Chapter 2

Cementing Basics
2.1 Introduction
Materials for cementing or bonding rock, brick, and stone in construction date from some of the
earliest civilizations. Remains of those early cements can still be found in Europe, Africa, the
Middle East, and the Far East. The calcareous (lime) cement that we know today had its
beginnings in the early gypsum and limestone cements.
Gypsum cements were used in the construction of the Egyptian pyramids approximately 3000
BC. These cements were made by calcining (heating) gypsum (CaSO4.2H2O) to approximately
260F to produce hemihydrate (CaSO4.0.5H2O). This hemihydrate, when rehydrated with a small
quantity of water, forms an interlocking mass of gypsum crystals that gives the cement its
strength. This process is still used to produce hemihydrate or, as it is better known, plaster of
Paris.
Limestone cements were first used around 800 BC in Crete, Cyprus, and Greece, and were
prepared by calcining limestone (CaCO3) at 1830F. The lime (CaO) produced was then mixed
with water, and hardening was essentially caused by the evaporation of the mix water. The
external surface of the hardened limestone was solidified further by subsequent carbonation with
CO2 from the air. The hardness and durability of the limestone cements was attributed not only to
the calcining process, but also to how the cement was mixed and placed.1
Gypsum and limestone cements are non-hydraulic. Even though they require water to react,
they do not harden under water. A hydraulic cement, on the other hand, is one that will harden
and gain strength even when stored under water. The earliest-known hydraulic cements were
used around 300 BC by both the Greeks and the Romans. These were produced by either
calcining limestones containing argillaceous (clayey) impurities or by mixing finely ground
materials such as potsherd, tiles, pottery, brick or volcanic tuffs with the lime. It was the silicates
in these added materials that reacted with lime to produce the hydraulic properties of these
cements. Santorin earth from the island of Thera, originally used by the Greeks to make cements
in 300 BC, is still used in construction in the Mediterranean today. The Romans used volcanic
tuffs from different points on or near the Bay of Naples. As the best variety of volcanic tuffs was
obtained from the neighborhood around the village of Pozzuoli, near Mt. Vesuvius, it was given
the name Pozzolana. The name pozzolana was then extended to cover the whole class of
materials having the same properties.2 Pozzolanic hydraulic cements were used not only in the
building of early Roman docks and marine facilities in the Mediterranean, but also Roman baths,
aqueducts such as the Pont du Gard at Nimes, the Appian Way and the Coliseum and Pantheon
in Rome. Pozzolanic materials are still used extensively as an extender in hydraulic cements and
for the improved durability they impart. (See Section 2.11.3.2)
Cementing technology started to decline after the Roman times and continued to deteriorate
throughout the Middle Ages (1200-1500 AD), during which time the art of burning lime was
almost completely lost. It was not until the late 1700s that cement technology started to advance.
John Smeaton was the first to recognize the factors that controlled the formation of hydraulic
lime when he was commissioned in 1756 to rebuild the Eddystone lighthouse, off the coast of
Cornwall. The hydraulic lime he used was prepared from limestone containing a high proportion
of argillaceous material and was mixed with a pozzolan imported from Italy. Then, in 1796,
James Parker patented the formation of hydraulic cement by calcining nodules of argillaceous

1
limestone known as septaria, found in certain Tertiary strata.3 A further development to this was
the calcining of mixtures of limestone and clay ground together in a wet mill, a process that may
be considered the precursor of Portland cement.4 In the early nineteenth century, numerous
patents for hydraulic cements were filed. The most historic of these was that of Joseph Aspdin,
an English mason, who was issued a patent5 covering a gray rock-like material called Portland
Cement dated 21 October 1824. The name Portland was chosen because of the resemblance of
the cement to Portland stonea limestone quarried in Dorset, Englandthat had a reputation for
quality and durability. The temperatures used to manufacture Aspdins early cements were not
sufficient to cause real clinkering, and the mineralogy and hydraulic activity were more
indicative of a hydraulic lime and not of the Portland cement currently manufactured.
Nevertheless, whatever the nature of Aspdins, cement it was the essential link in the
technological development of the Portland cement clinkers that we use today.
The concept of using hydraulic cement for oil wells was first reported in a patent by John R
Hill,6 though it was not until 1883 that Hardison and Stewart are reported to have put cement in a
well in Pico, California.7 The technological development in well cementing processes has
advanced considerably since those early days and continues to evolve as new challenges are
encountered.

2.2 Manufacture of Cement

Normal Portland cements are a manmade product that can be modified, depending on the raw
materials used and on changes to the process.8 From an overview, the manufacture of Portland
cement seems to be a relatively simple process. A calcareous material (calcium carbonate from
limestone or chalk) and an argillaceous material (aluminosilicate from clay or shale) are blended
together, wet, dry, or semi-dry, and fed into a rotary kiln. Rotary kilns range in length from 200
to 600 ft and have an average diameter of 12 to 16 ft. The mixture entering the kiln is
progressively heated to temperatures from 2550 to 2900F (1400 to 1600C). At these
temperatures, the material has the appearance of molten lava and it is in this partially melted
stage, the clinkering zone, that the calcium silicates, the active components of todays
hydraulic cements, are formed (Figure 2.1.). The cement clinker exiting the clinkering zone is
cooled rapidly. As the liquid cools, further reactions take place, with formation of calcium
aluminate and calcium aluminoferrite phases, and polymorphic transitions (changes in the crystal
structure) of the already formed crystalline calcium silicate phases. The cooled clinker is then
finely ground with a small amount of gypsum (about 5%) to form the gray powder sold as
Portland cement. Gypsum is added to retard the set of the calcium aluminate phase, which is
highly reactive in water. Although the process itself is simple, many variables affect the final
product. These can include trace elements in the raw mix, the degree of homogeneity of the raw
feed, particle size of the feed, the feed rate through the kiln, the kiln temperature, the source of
fuel, the rate of cooling, clinker storage, and grinding conditions. Good quality-control practices
are essential. Standards governing the quality of the cement produced are discussed in the
sections ahead. Oilwell cements should be consistent from one batch to another, and should give
reasonable results with additives typically used for downhole operations.
Proportioning of the raw materials is based on a series of simultaneous calculations that take
into consideration the chemical composition of the raw materials and the type of cement to be
produced: ASTM Type I, II, III, or V, white cement, or API Class A, C, G or H. In some cases,
additional or corrective components must be added to produce the optimum compositions.
Examples of such additives are sand, siliceous loams, pozzolans, diatomaceous earths, iron
pyrites, and alumina.9 Calculations also take into account argillaceous or siliceous materials that
may be present in high proportions in some limestones, or may result from the ash produced
when coal is used to fire the kiln. Minor impurities in the raw material also must be taken into
2
account, as they can significantly affect cement performance. For example, limestone deposits
can contain magnesite and minerals containing lead, zinc, copper, barium, fluoride, and
phosphate. Magnesia can cause delayed expansion in set cement, and the heavy metals, even in
low concentrations, can retard cement set. Quartz is often a major constituent in clays; clays may
also contain small amounts of calcite, gypsum, iron pyrites, organic matter, and alkalis derived
partly from exchangeable cations or from micas or feldspars present in the deposit. A small
proportion of the alkali incorporates into the crystal structure of the calcium silicates and calcium
aluminates formed during clinkering, but most of the alkali preferentially combines with sulfur
from the fuel or from gypsum in the clays to form alkali sulfates.10 These alkali sulfates, which
are deposited on the cement grain surfaces, are soluble in water, and are the source of the high
pH (>12) associated with cement slurries. Gelation and other detrimental effects in setting
cement have also been attributed to alkali sulfates. Iron is also a major substituent for silicon or
aluminum in clay minerals, and it forms a calcium alumino-ferrite phase in the cement clinker.
This phase is considered to be less reactive in setting cement than the corresponding calcium
aluminate phase, and is more sulfate-resistant.
High-quality cements require raw materials of suitable purity and uniform composition. This is
achieved, in part, by planned development of the deposit and stockpile blending before mixing
and proportioning the raw feed. Deposits found to contain too high a concentration of
detrimental minor phases are generally rejected.

2.3 Cement Nomenclature

Traditionally, cement compounds have been reported as the sum of the oxides, a notation
frequently used in ceramic chemistry. For example, the major component in cement, tricalcium
silicate, has the chemical formula Ca3SiO5 and in terms of the sum of the oxides is equivalent to
3CaO.SiO2. This has been further simplified in cement chemistry with a unique but useful
shorthand notation where a single capital letter replaces the oxide formulae. For example, C =
CaO: S = SiO2: A = Al2O3: F = Fe2O3: H = H2O: M = MgO: N =Na2O: K = K2O: P = P2O5: T =
TiO2. Thus, tricalcium silicate (Ca3SiO5) from its oxide equivalent 3CaO.SiO2 becomes simply
C3S.
In order to use simplified notation for carbonates and sulfates that are also important in cement
chemistry, the symbols C and S have been adopted, where C = CO2 and S = SO3. Calcium
carbonate (CaCO3), for example, can be written as CaO.CO2, which in shorthand notation
becomes C C . By the same token, gypsum (CaSO4.2H2O) can be written as CaO.SO3.2H2O,
hence C S H2. This shorthand convention was adopted to minimize space when writing more
complex formulas, and will be used along with conventional chemical notation throughout this
chapter.

2.4 Composition of Portland Cement

The composition of Portland cement is routinely determined by X-ray fluorescence (XRF) or
standard chemical methods, and the elements are defined in terms of the oxides. Trace oxides
such as TiO2, P2O5 and Mn2O3 are usually not given. This analysis is convenient from a quality-
control perspective, but it is a bulk analysis and does not give information on the actual phase
composition of a cement. In other words, cement is not a blend of individual oxides, but a
composite mixture consisting primarily of four major crystalline phases: tricalcium silicate
(C3S), dicalcium silicate (C2S), tricalcium aluminate (C3A), and a calcium aluminoferrite solid
solution known as the ferrite phase (Ca2(AlxFe1-x)2O5 where 0<x<0.7). The ferrite phase is more
commonly written as the tetracalcium aluminoferrite (C4AF), which is the midpoint in the solid

3
solution series where x = 0.5. A typical example of a mill-run analysis is listed in Table 2.1 for
both API Class G and H cements.
The phase composition relates to what class of cement is produced and its performance
characteristics on reaction with water (Table 2.2). Minor and trace elements considered as
impurities also play an important role in the reactivity of cements. They can substitute for ions in
the crystal structure of the cement phases, causing defects that can ultimately change the crystal
structure to produce different polymorphs (Table 2.3). Determining what polymorphs are
present, the number of defects, and how they affect cement performance is a complex task.11
Their effect on performance can be even more significant in the presence of oilwell additives,
and is a contributing factor to cement variability. In some cases, impurities are present in
quantities greater than can be incorporated into the crystal structures of the cement phases, and
are then observed as separate phases (e.g., alkali sulfates and periclase). Impurities can also
modify the chemistry of the reactions occurring during clinkerization, and can alter the
proportions of phases formed. An example of this would be increasing the amount of C3S with
respect to C2S, or vice versa. Impurities can also affect the crystal growth and ultimate crystal
size. In cement chemistry terminology, the terms alite, belite, aluminate (celite), and ferrite are
used to differentiate the phases actually found in cement that contain impurities from those of the
corresponding pure phases, C3S, C2S, C3A, and C4AF.
The quantitative phase composition can be determined from phase diagrams, but is generally
estimated using a simple calculation derived by Bogue12 and based on the oxide analysis and
compound stoichiometries. The calculation assumes that all the Fe2O3 occurs as C4AF solid
solution. The Al2O3 not included in the C4AF must then occur as C3A. C3S and C2S contents are
derived from two simultaneous equations for the SiO2 and CaO after CaO has been deducted for
C4AF, C3A, free lime, and lime present in the added gypsum. Depending on the Al2O3/Fe2O3
ratio, this leads to the equations given in Table 2.4.13 The Bogue calculation has inherent errors
in that it assumes the phases reach equilibrium during cooling, which does not necessarily occur.
It also considers that the phases are pure, and does not take into account the minor oxides that
can significantly change the composition and modify the crystal structure. Modified Bogue
calculations reported in the literature14 take into account the minor oxides, but these calculations
are not widely accepted.

2.5 Cement Analysis and Diagnostics

Several techniques are available for determining the characteristics of cement clinker or
powder, depending on what information is required. These techniques, although used primarily
for either quality control or research, are also applicable in troubleshooting field problems when
the integrity of the cement is in question. These techniques relate to the cement powder and
should not be used for investigating cement slurry or set cement. Techniques for investigating
slurries and set cements, including additives, will be discussed in Section 2.12.
As discussed previously, XRF or standard chemical analyses are used as routine quality-control
methods to determine the overall chemical composition of cement. For convenience, results are
given in terms of the oxides, and can be used in the Bogue calculation to give an estimate of the
phase (i.e., C3S, C2S, C3A, and C4AF).
Optical microscopy has become more prevalent over the years, not only for characterizing the
cement clinker, but also for determining the phase composition. The most widely used technique
has been to examine polished and etched sections of resin-impregnated samples in reflected
light.15 The etchants produce thin films of decomposition product that produce interference
colors in reflected light. The colors produced depend on the thickness of the reaction layer, and
vary according to the etchant used, the time of exposure to the etchant, and the reactivity of the
individual crystal. The phases can be identified by their characteristic shape (Figure 2.2).

4
Typically, alite and belite are found embedded in the aluminate and ferrite phases. Minor phases
are also often observed.
Alite (C3S). Angular or pseudohexagonal. Evidence of twinning, such as lamellae and Formatted: Indent: Left: 0.5", Bulleted +
Level: 1 + Aligned at: 0" + Tab after: 0.25" +
inclusions of belite, are often common in the internal structure. Average size of the Indent at: 0.25", Tab stops: 0.75", List tab +
crystals is 15 to 20m. Not at 0.25"
Belite (C2S). Rounded crystals, normally striated as a result of polymorphic
transformations. Average size of the crystals is 25 to 40m.
Aluminate (C3A). The cubic form has no definite structure and tends to fill in the
spaces between other phases, whereas the orthorhombic form produces elongated
prismatic crystals. The orthorhombic form is produced if a substantial amount of
alkali oxide is incorporated in the structure.
Ferrite (C4AF). Tends to be in intimate contact with the aluminate phase, and is
differentiated by its lighter color. Often forms dendritic or needles in the aluminate
matrix.
Periclase (MgO). Forms in clusters and has a well-formed structure.
Free lime (CaO). Smooth, rounded grains, usually in clusters and surrounded by alite
crystals. Typical size is 10 to 20m.
Alkali sulfates. Occur in the clinker pore structure. They can be observed as black
particles with a hydrofluoric acid etch or as empty pores if treated with an aqueous
etch.
Quantitative determination of the phase composition is determined using a point-counting
technique, ASTM C1356 M-98.24 The phase observed at each point is noted, with some 2,000 to
4,000 counts considered necessary to give reasonable precision.16 The areas of the phases in the
polished section are proportional to the volumes in the material, and can be converted to weight
percentages using individual phase densities. Differentiation of the aluminate and ferrite phase is
often difficult because of the small size of the crystals. Other inherent errors in this technique are
associated with imperfect sampling, size of the grid, number of points analyzed, the grain sizes,
and incorrect phase identification.
QXRD (quantitative X-ray diffraction) has also been used over the years to determine the
phase composition. This technique produces significantly more information on the phases, such
as polymorphic modifications, the state of crystallization, and other features of the individual
phases that all relate to the effect of the minor components in the C3S, C2S, C3A, and C4AF
crystal structure. Quantitative analysis used to be time-consuming, and was generally the domain
of researchers. The problems with QXRD were preferred orientation in some of the phases and
the wide range of overlapping reflections of the major components. Selective dissolution
techniques that preferentially dissolve the alite and belite17 or the aluminate and ferrite phases18
are often coupled with QXRD analysis (Figure 2.3). An internal standard method using single-
peak measurements was good for the aluminate and ferrite phases, but not for the silicates,
particularly belite, where the only peak free from overlapping was of low intensity. ASTM
C1365-98 gives a standard method for using QXRD for Portland cement and Portland cement
clinker. Advances in electronic data processing have led to the Rietveld analysis that calculates,
compares, and refines X-ray diffraction patterns from complex-structure data and can analyze
phase mixtures of up to 15 components. This method is reported19 to give accurate data in a short
time, for both the major phases and for minor phases present above 0.3 wt%.
Optical microscopy in conjunction with QXRD is considered to give the best information on
the phase composition, though presently the Bogue calculation is still used by most
manufacturers (Table 2.1). Optical microscopy is used more often as a diagnostic tool because it
produces much information in a short time and allows for changes to be made during the

5
manufacturing process so that good clinker is obtained. For example, clusters of alite or belite
suggest that the raw mix is not homogenous, while small belite crystals formed as a fringe on
alite crystals indicate a slow cooling rate. Optical microscopy is typically performed on cement
clinker, though microstructural and interpretation can be transferred to the ground cement
powder but the small size of the particles makes large-scale systematic investigations more
difficult.20
Clearly, there is still much to be done toward understanding the chemical nature and physical
properties of cement and how these properties affect performance. Currently, ASTM and API
procedures use performance tests that are appropriate for field applications as part of the quality-
control procedure and as a means of obtaining a consistent product for application. API
procedures include thickening time, consistency, compressive strength, and free fluid (Tables
2.1 and 2.5.)
The variability of cement is a major issue, particularly in slurry design where additives are
almost always used. Significant differences can occur in using different sources of cement and,
to a lesser extent, among batches of cement from the same source. Attempts have been made
over the years to correlate slurry performance with cement characteristics.21-23 These attempts
have shown certain correlations, such as greater reactivity with increased fineness or alite crystal
size, reduced thickening time with formation or presence of syngenite, greater expansion with
increased free lime or periclase content, etc. Such cement characteristics tend to be more
predominant, and their effect on performance is more readily observed. However, these
correlations hold true only when all other factors are equal, and from one source of cement to
another, they rarely are. There are so many variables that can enhance and/or counteract the
reactivity of one or more crystalline phases that no method or compilation of methods to date can
fully characterize the cement, much less predict its performance.

2.6 Classification of Cement

Portland cements are usually manufactured to meet certain chemical and physical standards
that depend upon their application. In the U.S., several agencies study and write specifications
for the manufacture of Portland cement. These groups include American Concrete Institute
(ACI), American Association of State Highway and Transportation Officials (AASHTO),
American Society for Testing Materials (ASTM), American Petroleum Institute (API), and
various departments of the federal government. Of these groups, the best known to the oil
industry are the ASTM, which deals with cements for construction and building use24, and the
API, which writes specifications for cements used only in wells.13,25 Cement specifications writ-
ten by either society are prepared by representatives of both users and manufacturers working
together for the common interest of their industry. Specifications govern both the chemical
properties of the cement and the physical requirements for determining the performance.
The ASTM specification C150 provides for eight types of Portland cement: Types I, IA, II,
IIA, III, IIIA, IV, and V, where the A denotes an air-entraining cement.24 These are designed to
meet the varying needs of the construction industry, from building blocks to dams. Cements used
in wells are subjected to conditions not encountered in construction, such as wide ranges in
temperature and pressure. For these reasons, different specifications were designed and are
covered by API Specifications.13 API currently provides specifications for eight classes of
oilwell cements, designated Classes A through H. Initially, the performance of cement was
controlled by modifying the phase composition of the cement (Table 2.2), but now additives are
used to better control slurry designs for specific application. Because of this, the number of API
classes has been reduced to the point that API Classes G and H are the most widely used. In
Europe, Canada, the Middle East, South America, and the Far East, Class G cement is used most
frequently. In the U.S., about 15% of API Class G cement is used in California and the Rocky

6
Mountain areas, while approximately 65% of the API Class H cement is used in Gulf Coast and
mid-continent operations. The other classes of cement used in cementing wells are typically API
Class A (10%) and API Class C (10%). API Class B is used occasionally. Specialty cements
(Section 2.9) constitute less than 1% of the worldwide market.
Oilwell cements are also available in either moderate sulfate-resistant (MSR) or high sulfate-
resistant (HSR) grades. Sulfate-resistant grades are used to prevent deterioration of set cement
downhole caused by sulfate attack by formation waters. In sulfate-resistant cement grades, the
amount of C3A produced is minimized by adding Fe2O3 to the raw feed during manufacture to
produce more ferrite phase.

API Classifications. The oil industry purchases cements manufactured predominantly in

accordance with API classifications as published in API Standards 10A, Specifications for
Cements and Materials for Well Cementing.13 These standards have been published annually by
the American Petroleum Institute since 1954, when the first national standards on cements for
use in wells were issued. These specifications are reviewed annually and revised according to the
needs of the oil industry. The different classes of API cements for use at downhole temperatures
and pressures are defined in the following list. They are listed* in the API Standards 10A, dated
January 1, 1995.13

Class A: The product obtained by grinding Portland cement clinker, consisting essentially of
hydraulic calcium silicates, usually containing one or more of the forms of calcium sulfate as an
interground material. At the option of the manufacturer, processing additions may be used in the
manufacture of the cement, provided such materials in the amounts used have been shown to
meet the requirements of ASTM C 465. This product is intended for use when special properties
are not required. Available only in ordinary (O) Grade (similar to ASTM C 150, Type I).**

Class B: The product obtained by grinding Portland cement clinker, consisting essentially of
hydraulic calcium silicates, usually containing one or more of the forms of calcium sulfate as an
interground material. At the option of the manufacturer, processing additions may be used in the
manufacture of the cement, provided such materials in the amounts used have been shown to
meet the requirements of ASTM C 465. This product is intended for use when conditions require
moderate or high sulfate resistance. Available in both moderate sulfate-resistant (MSR) and high
sulfate-resistant (HSR) grades (similar to ASTM C 150, Type II).

*API Specification 10A: Specification for Cements and Materials for Well Cementing.
Copies of this specification are available from American Petroleum Institute, 1220L
Street, Northwest, Washington, D.C. 20005: website www.api.org
** ASTM C 150: Standard Specification for Portland Cement. Copies of this specification
are available from American Society for Testing Materials, 100 Bar Harbor Drive, P.O.
Box C700, West Conshohocken, Pennsylvania 19428-2959: website www.astm.org

A suitable processing addition or set-modifying agent shall not prevent a well from
performing its intended functions. 1

Class C: The product obtained by grinding Portland cement clinker, consisting essentially of
hydraulic calcium silicates, usually containing one or more of the forms of calcium sulfate as an
interground material. At the option of the manufacturer, processing additions may be used in the

7
manufacture of the cement, provided such materials in the amounts used have been shown to
meet the requirements of ASTM C 465. This product is intended for use when conditions require
high early strength. Available in ordinary (O), moderate sulfate-resistant (MSR), and high
sulfate-resistant (HSR) grades (similar to ASTM C 150, Type III).

Class G: The product obtained by grinding Portland cement clinker, consisting essentially of
hydraulic calcium silicates, usually containing one or more of the forms of calcium sulfate as an
interground addition. No additions other than calcium sulfate or water, or both, shall be
interground or blended with the clinker during manufacture of Class G well cement. This product
is intended for use as a basic well cement. Available in moderate sulfate-resistant (MSR) and
high sulfate-resistant (HSR) grades.

Class H: The product obtained by grinding Portland cement clinker, consisting essentially of
hydraulic calcium silicates, usually containing one or more of the forms of calcium sulfate as an
interground addition. No additions other than calcium sulfate or water, or both, shall be
interground or blended with the clinker during manufacture of Class H well cement. This product
is intended for use as a basic well cement. Available in moderate sulfate-resistant (MSR) and
high sulfate-resistant (HSR) grades.

Surface areas based on the Blaine method for Class G and Class H cements typically lie in the
range of 300 to 400 and 220 to 330 m2/kg, respectively.

2.7 Properties of Cement Covered by API Specifications

In well-completion operations, cements are almost universally used to displace the drilling mud
and to fill the annular space between the casing and the open hole. To serve this purpose,
cements must be designed for wellbore environments varying from surface conditions to
conditions at depths exceeding 30,000 ft, where temperatures range from below freezing in
permafrost areas to more than 700F in wells drilled for geothermal steam production. No
published specifications for blends of well cements or optional additives cover properties of ce-
ments over such broad ranges of depth and pressure. API Standards 10A specifications do,
however, give chemical and physical data for the different classes of cements required for use in
well conditions.13 API specifications are not enforced by an official agency; however, use of the
API monogram indicates that the manufacturer has agreed to make cement according to the
specifications outlined in the API Standards 10A and API Specification Q1. Chemical properties
and physical requirements, as defined by API Standards 10A, are summarized in Tables 2.4 and
2.5. Typical physical requirements of the various API classes of cement are shown in Table
2.6.26 Although these properties describe cements for specification purposes, oilwell cements
should have other properties and characteristics to provide for their necessary functions
downhole.27 API Recommended Practice 10B provides standards for testing procedures and
special apparatus used for testing oilwell cements and includes
slurry preparation
slurry density
compressive-strength tests and non-destructive sonic testing
thickening-time tests
static fluid-loss test
permeability tests
rheological properties and gel strength

8
 pressure drop and flow regime calculations for slurries in pipes and annuli
arctic (permafrost) testing procedures
slurry stability test

compatibility of wellbore fluids25

Originally, API determined average well conditions from well data obtained mostly from the
Gulf of Mexico and the southwestern United States, and used these data to develop standard
operation schedules. Depending on the thermal gradient, wells of equal depth in different
geographical locations can have different bottomhole temperatures. As a result, new pressure-
temperature schedules have been included in API Recommended Practice 10B, though with
variable thermal gradients in locations around the world, it is not possible to cover all well
conditions. As such, the use of the pressure-temperature schedules and other information in API
Recommended Practice 10B is optional and the information can be modified to represent actual
well conditions.

2.8 Cement Standards Outside the U.S.

The API Standards 10A specification is the most widely used oilwell cement specification
worldwide, though several countries have developed their own standards28. Some of these have
been developed independently, as in China (GB202-78) and Hungary (MSZ 4702/8-82), while
others are based on a version of API specifications: CIS (formerly the USSR)(GOST 1581-96),
India (IS:8229E-1976), Indonesia (S11, 2204), Argentina (IRAM 1518-87), Brazil (NBR 9831-
93), Romania (STAS 1455-81:STAS 9253-90). In China, API Class G or H is increasingly
manufactured for well-cementing operations, whereas in Hungary there is a move toward
European Standards. The extent to which these country standards are used varies. In some
countries, it may be necessary to use local products that are manufactured for the construction
industry. Table 2.7 lists classifications that have been established for the most common types of
Portland cement in countries commonly associated with the oil industry.29 A specific
manufacturer may, for speed and simplicity, use a symbol to identify various types of cement.
Expansion of the global economy has led to an increased need for international standards for
materials, products, processes and services. As early as 1947, a nongovernmental organization,
the International Organization for Standardization (ISO), was established in Geneva,
Switzerland, to facilitate the international coordination and unification of industrial standards. To
date, ISO has a worldwide federation of national standard bodies from some 130 countries and
several hundred thousand standards and technical regulations in use throughout the world.30,31
Having an ISO Standard approval gives the standard complete international credibility and
acceptance. The API is currently working with ISO to have API specifications become ISO
standards. API standards for cement have been short-tracked to ISO standards with the goal of
making ISO and API standards consistent. To date, the ISO equivalent of API Specification 10A,
ISO Standard 10426-1, is the only one fully revised and approved.* ISO Standard 10426-2 which
is equivalent to API Specification 10B is in the formal approval stage and should be available in
2003.

Footnote
* ISO 10426-1: copies of this specification are available from American Petroleum institute,
1220L Street Northwest, Washington, D.C. 20005: website www.api.org

2.9 Specialty Cements

A number of cementitious materials that have been used very effectively for cementing wells do
not fall into any specific API or ASTM classification. While these materials may or may not be
9
sold under a recognized specification, their quality and uniformity are generally controlled by the
supplier. These materials include
pozzolanic-Portland cements
pozzolan-lime cements
resin or plastic cements
gypsum cements
microfine cements
expanding cements
refractory cements
latex cements
cements for permafrost environments
Sorel cements

2.9.1 Pozzolanic Cements

Pozzolanic materials include any natural or industrial siliceous or silico-aluminous material,
which, though not cementitious in itself, will combine with lime in the presence of water at
ambient temperatures to produce strength-developing insoluble compounds similar to those
formed from hydration of Portland cement.33 Typically, pozzolanic materials are categorized as
natural or artificial, and can be either processed or unprocessed. The most common sources of
natural pozzolanic materials are volcanic materials and diatomaceous earths. Diatomaceous
earths are composed of diatom fossil remains consisting of opaline silica. Artificial pozzolanic
materials are produced by partially calcining natural materials such as clays, shales, and certain
siliceous rocks, or are more usually obtained as an industrial byproduct. Artificial pozzolanic
materials include metakaolin, fly ash, microsilica (silica fume), and ground granulated blast-
furnace slag.
Metakaolin is formed by calcining china clay that contains the mineral kaolinite to a
temperature between 1110 and 1470F. Metakaolin is a very reactive pozzolan, and virtually
consumes all the lime available in a cement blend. It is used in well cementing at temperatures
below 60F.34
Fly ash is a combustion byproduct of coal, and is the most widely used pozzolan in the oil
industry. The chemical composition depends primarily on the coal sourcelignite, sub-
bituminous or bituminousand to a lesser extent on the mineral composition of the inorganic
component in the coal (gangue).35 It is the only pozzolan covered by both API and ASTM
specifications.
Microsilica is produced when quartz is reduced with coal and iron or other ores in electric arc
furnaces where temperatures approach 3600F. The silicon gas produced is oxidized by air and
condenses as amorphous silica containing particles in sizes ranging from 0.1 to 0.2 m. Although
the microsilica is pozzolanic, it is not normally used as a pozzolanic cement; its use in oilwell
cementing is determined by the other properties it exhibits.
Granulated blast furnace slag is a byproduct of the iron industry and is formed when iron
oxide ore is reduced to metallic iron in a blast furnace at temperatures of 2550 to 2730F. Widely
different slag products are obtained, depending on the process used to cool the slag. The
granulated blast furnace slag is formed on rapid cooling, producing a solidified glass that is
subsequently ground.
Pozzolanic oilwell cements are typically used to produce lightweight slurries. The specific
gravity of the pozzolanic material is lower than that of the cement and results in a pozzolan
slurry with a lighter weight than a corresponding Portland cement slurry of similar consistency.

10
Some pozzolanic materials also have a high water requirement that produces a higher yield and
lighter slurry. (See Section 2.11.3.2)
The pozzolanic materials tend to improve compressive strength over time because of the
reduction of Ca(OH)2 and increase in formation of binding material. The effectiveness of the
different pozzolanic materials is illustrated in Table 2.8. The additional binding material also
reduces permeability and minimizes attack from formation waters. In most cases, pozzolanic
materials can also reduce the effect of sulfate attack, though this is somewhat dependent on the
slurry design.36
Commercial cements such as TXI Lightweight (Trademark of TXI) for use in oil wells are a
special formulation composed of Portland cement clinker interground with lightweight siliceous
aggregate to produce a pozzolanic-like cement.

2.9.2 Pozzolan-Lime Cements

Pozzolan-lime or silica-lime cements are usually blends of fly ash (silica), hydrated lime, and
small quantities of calcium chloride.26,37 These products hydrate with water to produce calcium
silicate hydrates, calcium aluminate hydrates, and calcium aluminosilicate hydrates. The
pozzolanic activity of such cements depends on the maximum amount of lime that the pozzolan
can react with and the rate of the reaction. The nature and content of the active phase, the
lime/pozzolan ratio, length of curing, water-to-solids-mix ratio, and temperature all impact on
the pozzolanic activity of the pozzolan-lime cement. At low temperatures, the initial reactions of
these cements are slower than similar reactions in Portland cements. Therefore, they are
generally recommended for primary cementing at temperatures above 140F (284C). Fly ash is
composed primarily of an amorphous silicate or calcium aluminosilicate glass containing a small
amount of crystalline minerals. This glass dissolves in highly alkaline solutions produced from
the hydrated lime and precipitates on reaction with Ca2+ from the fly ash and the lime to produce
the hydration products.
The merits of this type of cement are ease of retardation, light weight, economy, and strength
stability at high temperatures.

2.9.3 Gypsum Cements

Gypsum cement is a blended cement composed of API Class A, C, G or H cement and the
hemihydrate form of gypsum (CaSO4 .1/2H2O). In practice, the term gypsum cements normally
indicates blends containing 20% or more gypsum. Gypsum cements are commonly used in low-
temperature applications for primary casing or remedial cementing work. This combination is
particularly useful in shallow wells to minimize fall-back after placement. The unique properties
of gypsum cement are its capacity to set rapidly, its high early strength, and its positive
expansion (approximately 2.0 %). This is caused by the plaster of Paris reaction where the
hemihydrate rehydrates to form gypsum.
CaSO4 .1/2H20 + 1.5H2O CaSO4.2H2O

A cement with high gypsum content has increased ductility and acid solubility. It is usually used
in situations of high lateral stress or in temporary plugging applications. A 50:50 gypsum cement
is frequently used in fighting lost circulation, to form a permanent insoluble plug. These blends
should be used cautiously because they have very rapid setting properties and could set
prematurely during placement. A limitation of gypsum cements is that they are non-hydraulic
(Section 2.1) and they are not stable in contact with external water sources.

11
2.9.4 Microfine Cements
Microfine cements are composed of very finely ground (1) sulfate-resisting Portland cements, (2)
Portland cement blends with ground granulated blast furnace slag and (3) alkali-activated ground
granulated blast furnace slag. Specific surface area for microfine cements is 500 to 1000 m2/kg
and sometimes higher. Microfine cements have an average particle size of 4 to 6 microns, and a
maximum particle size of 15 microns. They hydrate in the same manner as normal Portland
cements, though at a significantly faster rate because of the greater surface area. (See Section
2.10). The blends of Portland cement and ground granulated blast furnace slag cement are
equivalent to a finely ground pozzolanic cement, resulting in a faster hydration reaction. (See
Sections 2.9.1 and 2.11.3.2). Such cements have a high penetrability and ultra-rapid hardening.38
Applications for such cements include consolidation of unsound formations and repair of casing
leaks in squeeze operations, particularly tight leaks that are inaccessible by conventional
cement slurries because of their penetrability.39-40 Ultrafine alkali-activated ground blast furnace
slag is used in the mud-to-cement technology in which water-based drilling mud is converted to
cement. 41-42

2.9.5 Expanding Cements

Expansive cements are available primarily for improving the bond of cement to pipe and
formation. If expansion is properly restrained, its magnitude will be reduced and a prestress will
develop. Expansion can also be used to compensate for shrinkage in neat Portland cement.
Expansive cements were developed in the 1950s and 1960s that are hydraulic in nature and have
controlled expansion that occurs just after setting.43-45 These cements were based on either the
formation of considerable quantities of ettringite (C6A S 3 .H32) after set, or on hydration of
anhydrous polyvalent metal oxides such as MgO or hard-burnt CaO. In the late 1970s, in-situ
gas-generating additives were developed. These additives produce micro-size gas bubbles that
cause the cement to expand while still in the plastic state.

ASTM C845-96 covers three types of ettringite-expansive cement: Type K, Type S, and Type
M. Type S and Type M are no longer commercially produced in the United States. Type O
cement, not covered by ASTM specifications, is a hard-burnt CaO expansive cement, originally
produced in Japan.

Type K cement contains the anhydrous calcium sulfoaluminate C4A3 S , which can be formed in
the rotary kiln as an integral part of the cement or formed separately and then blended with
Portland cement and anhydrite C S . Most Type K cements also contain some uncombined CaO
to change the early rate of expansion. The stoichiometric expansion reaction is:
C4A3 S + 8 C S + 6 CH + H 3 C6A S 3 .H3

The CH is obtained on hydration of the Portland cement. Expansion from the reaction in the set
cement is approximately 0.05 to 0.20%.

Type S cement, suggested by the Portland Cement Association, is a Portland cement that is
similar to API Class A, but has a very high C3A content (approximately 20%). The cement also
contains 10 to 15% gypsum. Expansion characteristics are similar to those of Type K cement,
though the reaction is somewhat different:
C3A + 3 C S + H C6A S 3 .H32

12
Type M cement is obtained by adding small quantities of refractory cement (with calcium
aluminate as the principal compound ) and additional calcium sulfate to Portland cement to
produce expansive forces.
CA + 3 C S + 2 CH + H C6A S 3 .H32

Type O cement, developed by the Onoda Cement Company, derives its expansion properties
from the delayed hydration of the hard-burnt CaO:
CaO + H Ca(OH)2

Other formulations of expanding cement include the following:

API Class A (Portland cement) containing 5 to 10% of the hemihydrate forms of
gypsum.46 (The expansion characteristics of API Class A and Class H cements containing
gypsumcalcium sulfateare compared in Table 2.9.)
API Class A, G, or H cements containing sodium chloride in concentrations ranging from
5% to saturation. The expansion is caused by calcium chloroaluminate reactions (See
Section 2.11.1).

In-situ gas generation is based primarily on the reaction of finely ground alumina powder with
the alkalis present in the cement aqueous solution to produce hydrogen gas:

2Al(s) + 2OH-(aq) + 6H2O 2Al(OH)4-(aq) + 3H2(g)

Although alumina powder is the most commonly used, zinc, magnesium, and iron powders are
potential alternatives.

At this time, the API standards contain no test procedure or specifications for
measuring the expansion forces in cement. Most laboratories use the expansive
bar test, using a molded 1- 1- 10-in. cement specimen. Ring molds are also
available, though they are less commonly used. The expansive force is measured
shortly after the cement sets for a base reference and then at various time
intervals until maximum expansion is reached. Hydraulic bonding tests have also
been used to evaluate cement expansion.47

2.9.6 Calcium Aluminate Cements

High-alumina cement (HAC) was first developed for industrial use as a solution to degradation
of mortars and concretes in ground containing large quantities of sulfate.48,49 Today, these
cements are primarily used in refractory concretes, but they are also widely used in construction
for rapid setting and controlled expansion or shrinkage compensation. In well-cementing
operations, they are used at both temperature extremes in permafrost zones with temperatures at
32F or below; in-situ combustion wells (fireflood) where temperatures may range from 750 to
2000F, and thermal recovery wells where temperatures can exceed 1300F and fluctuate
dramatically.

Several high-alumina cements have been developed with alumina contents of 35 to 90 percent,
and there is a move to term these collectively as calcium aluminate cements (CAC) because the
reactive phase in all cases is calcium aluminate. Calcium aluminate cement is manufactured by
blending bauxite (aluminum ore) and limestone and heating the mixture above 2640F in

13
reverberatory open hearth furnaces until it is liquefied.50,51 The molten clinker is continuously
removed through a tap hole, collected in molds, cooled, and ground in ball mills. The setting
time for calcium aluminate cement is controlled by the composition and no materials are added
during grinding.

The manufacturer usually controls standards for calcium aluminate cements because few national
standards address these cements. Some typical composition ranges for the calcium aluminate
cements are listed in Table 2.10, in terms of the oxides for the major components. Like Portland
cement, the calcium aluminate cements are polymineralic and contain up to five phases:
monocalcium aluminate (CA), dodecacalcium heptaaluminate (C12A7), pleochroite
(Ca20Al22.6Fe(III)2.4Mg3.2Fe(II)0.3Si3.5O68), gehlenite (Ca2Al2SiO7), and ferrite solid solution.

The phases primary involved on hydration of CAC at early ages are CA and C12A7, the other
phases being relatively inactive. At temperatures above 86F (30C), the ferrite solid solution
also plays a significant role in the hydration. The hydration reaction is considered to be a
through-solution reaction where the solid anhydrous phases first dissolve and then the hydration
products precipitate from the solution. This process is slow at first as the hydration products
nucleate and grow; then, it accelerates. When the hydration products attain a critical size,
massive precipitation occurs and the solution concentrations drop. This is associated with setting
and occurs after about 2 hours. The initial hydration products formed are strongly dependent on
temperature, and are:
CAH10 - < 50F (10C)
CAH10 + C2AH8 - 32 to 80.6F (0-27C)
C3AH6 + AH3 - > 80.6F (27C)

The CAH10 and C2AH8 are metastable phases and ultimately will convert to the stable
hydrogarnet (C3AH6) and gibbsite (-AH3) forms. This process can take years at low
temperatures or it can occur immediately at high temperatures. An amorphous or poorly
crystalline gel phase based on hydrous alumina (AH3 gel) may also form along with the
crystalline products.

It is the standard type (e.g., Ciment Fondu) that is most commonly used in well cementing. These
cements can be accelerated or retarded to fit individual well conditions, however, the retardation
characteristics differ from those of Portland cements. The addition of Portland cement to a
refractory cement will cause a flash set; therefore, when both are handled in the field, they must
be stored separately.

The annual production of calcium aluminum cements is very small by comparison with Portland
cement, and they are considerably more expensive. Calcium aluminate cements are
manufactured worldwide and can be obtained in the United States, UK, France, Spain, Eastern
Europe, Brazil, China, Japan, Korea, India, and CIS (formerly the USSR).

2.9.7 Latex Cement

Latex cement, although sometimes identified as a special cement, is actually a blend of API
Class A, G, or H with latex. Latex is a colloidal suspension of polymer in water. Latex in latex
cements are generally copolymer systems that incorporate more than one type of polymer to
optimize film formation and flexibility. The copolymers are based on polyvinyl acetate,
polyvinylidene chloride-polyvinyl chloride, polyacrylate copolymers or styrene-butadiene, and
14
are spherical with diameters of 0.01 to 1.0 micrometers. In general, a latex emulsion contains
only 50 percent by weight of solids and is usually stabilized by an emulsifying surface-active
agent. The latex particles coalesce to form a continuous film around the cement hydration
products in the set cement and effectively coat the walls of the capillary pores. Latex imparts
elasticity to the set cement and improves the bonding strength and filtration control of the cement
slurry. In powdered form, latex is not susceptible to freezing and can be dry-blended with the
cement before it is transported to the wellsite.

2.9.8 Permafrost Cement

Special problems can occur in cementing conductor and surface casing in a frozen
environment.52 Throughout the Arctic, ice-bearing formations extend to depths up to 3,000 ft.
They may be described as frozen earth or glacier-like ice blocks.53,54 Operators typically prefer to
use a quick-setting, low-heat-of-hydration cement that will not melt the permafrost. API
Recommended Practice 10B, Section 14, gives special cementing procedures for simulating
Arctic conditions and cementing in such environments.25

Two cement systems that have been used successfully are calcium aluminate cement blends and
gypsum cement blends.52,56 Fly ash or natural pozzolan is normally blended (approximately 50%
by weight) with calcium aluminate cements to lower the heat evolved during hydration, thus
preventing permafrost damage.57 Gypsum-cement blends can be accelerated or retarded and will
set at 15F below freezing. For surface pipe, these slurries are normally designed for 2- to 4-hour
pumpability, yet their strength development is quite rapid and varies little at temperatures
between 20 and 80F.

2.9.9 Resin or Plastic Cements

Resin and plastic cements are specialty materials used for selectively plugging open holes,
squeezing perforations, and cementing waste disposal wells, especially in highly aggressive,
acidic environments. They are usually mixtures of water, liquid resins, and a catalyst blended
with an API Class A, B, G, or H cement. For example:

A unique property of these cements is their ability to be squeezed under applied pressure into a
permeable zone to form a seal within the formation. These specialty cements are used in
relatively small volumes, and are effective at temperatures from 60 to 200F (140 to 392C).

2.9.10 Sorel Cement

Sorel cement is a magnesium-oxychloride cement used as a plugging material in well cementing.
The cement is made by mixing powdered magnesium oxide with a concentrated solution of
magnesium chloride. The complex hydration reactions include at least eight different primary
reactions.33 Carbonates are generally incorporated into the formulation to reduce the solubility of
the magnesium hydroxide chloride hydrates that are normally formed by producing carbonated
hydrates. The main phases formed are Mg2OHClCO3.3H2O and Mg3(OH)2(CO3)4.4H2O.59 Sorel
cements have been used on occasion in the CIS for cementing oil wells at temperatures up to
15
1400F (752C).56 An acid-soluble magnesia cement that reacts as a complex Sorel cement has
been set across production perforations as a temporary abandonment plug and used to protect
water injection zones during workover operations.60

2.10 Cement Hydration

The reactions involved when cement is mixed with water are complex. Each phase hydrates by a
different reaction mechanism and at different rates (Figure 2.4). The reactions, however, are not
independent of each other because of the composite nature of the cement particle and proximity
of the phases. (See Fig 2.2 and Section 2.4). Only very small particles, measuring 1m or less,
tend to exist as individual phases. Cement hydration occurs in a sequence of stages as illustrated
in Figure 2.5. In all, five distinct stages have been identified: (1) pre-induction, (2) dormant
(induction) period, (3) acceleration, (4) deceleration and (5) steady state. The chemical processes
occuring at each stage are summarized in Table 2.11.

In cementing operations, Stages 1 through 3 are of greatest importance. Stage 1 dictates the
initial mixability of the cement and is attributed primarily to the aluminate and ferrite phase
reactions. Stage 2 relates to the pumpability time, and Stage 3 indicates setting properties and gel
strength development. This reaction sequence poses several questions from a chemistry
perspective:
Why is there a dormant period?
What stops the initial reaction from proceeding until the cement is completely reacted?
Why does the dormant period end and the reaction take off again?
Understanding these processes is beneficial for modifying cement slurry designs and for
determining suitable additives for well cement applications.

The literature is abundant with studies on both laboratory-synthesized pure phases (C3S, -C2S,
C3A, C4AF) and cement, but these studies contain numerous and often conflicting data. This is in
part due to the complexity of the reaction, even for the pure phases, and in part due to the lack of
a single technique that will provide all the necessary information. A major obstacle is that the
principal hydration product, calcium silicate hydrate (C-S-H), formed on reaction of the alite and
belite silicates with water, is an amorphous gel with almost no defined structure (Figure 2.6). It
is composed of particles <1m in size and its composition changes with time and temperature.
The C-S-H gel formed on hydration constitutes about 70% of the fully hydrated cement, and
gives the cement its rigidity and compressive strength.

2.10.1 Hydration of Pure Mineral Phases

C3S. On reaction with water, C3S produces C-S-H and calcium hydroxide, CH, (also known as
Portlandite). The hyphens used in the C-S-H formula indicate its variable composition: CSH
would imply a fixed composition of CaO.SiO2.H2O. C/S ratios in C-S-H vary from 1.2 to 2.0,
and H/S ratios vary from 1.3 to 2.1.

2C3S + 6H C-S-H + 3 CH

In contrast, the CH formed during the reaction is a crystalline product with a fixed composition,
Ca(OH)2. CH imparts no intrinsic properties to the set cement. The microstructural development
of C-S-H and CH are illustrated in Figure 2.7.

16
Over the years, several theories have been postulated to explain why the initial reaction slows
minutes after mixing and why the reaction is renewed after the so-called dormant or
pumpability period:
o1. Formation of an electrical double layer Formatted: Indent: Left: 0.5", Outline
numbered + Level: 1 + Numbering Style: 1, 2,
Incongruent dissolution of C3S creates a silicate-rich surface layer with adsorbed 3, + Start at: 1 + Alignment: Left + Aligned
calcium ions, producing an electric double layer that slows down movement of ions at: 0" + Tab after: 0.25" + Indent at: 0.25",
into solution, giving rise to the onset of the dormant period. Termination of the Tab stops: 0.75", List tab + Not at 0.25" + 1"

dormant period and renewed hydration occurs on the breakdown of this double layer.
61

2. Replacement of a protective first hydrate by a second, more permeable hydrate

The dormant period commences with the formation of a protective semi-permeable
gelatinous C-S-H coating around the C3S grain surface and slows down movement of
water and/or ions to or from the C3S grain surface and the aqueous phase.62-67 The end
of the dormant period occurs
a) when water diffuses through the semi-permeable C-S-H coating and causes an
osmotic pressure to build up between the C-S-H coating and the C3S grain,
eventually causing the coating to rupture68,69 or
b) on formation of a second-hydrate C-S-H gel, due to phase transformation or
aging effects, that is more permeable to water and ions.70,71
3. Nucleation and growth of C-S-H
The rate of reaction during the dormant period and the renewed hydration of C3S are
controlled by nucleation and growth of a second-hydrate C-S-H that differs from the
first hydrate formed. This second-hydrate C-S-H forms after the thermodynamic
barrier for its formation has been overcome, producing large numbers of nuclei and
rapid growth. The dormant period ends when the nucleation and growth of the
second-hydrate C-S-H begins. 72-76
4. Nucleation of calcium hydroxide
The dormant period is considered to be a classical induction period, dependent on
supersaturation in this case, with respect to calcium hydroxide. Growth of calcium
hydroxide nuclei is poisoned by silicate ions and the induction period ends when the
level of supersaturation is sufficient to overcome this.61,66,77,78

None of these theories totally explains all the phenomena observed. The theories are, however,
more complementary than contradictory, and the true reaction mechanism is most likely a
combination of these theories.

C2S. The kinetics and hydration mechanism for C2S are similar to those of C3S except that the
rate of reaction is much slower. The hydration products are the same except that the proportion
of calcium hydroxide produced is about one-third of that obtained on hydration of C3S.

2C2S + 4H C-S-H + CH

C3A . The initial reaction of C3A with water in the absence of gypsum is vigorous and can lead
to flash set caused by the rapid production of the hexagonal crystal phases, C2AH8 and C4AH19,
(Figure 2.8). Sufficient strength is developed to prevent continued mixing. The C2AH8 and
C4AH19 subsequently convert to cubic C3AH6 (hydrogarnet), which is the thermodynamically
stable phase at ambient temperature. 79-81 Typically, gypsum is added to retard this reaction,
though other chemical additives can be used.

17
The products formed on reaction of C3A in the presence of gypsum depend primarily on the
availability of sulfate ions from the dissolution of gypsum (Figure 2.9). The primary phase
formed is 6-calcium aluminate trisulfate-32-hydrate, commonly known as trisulfate or ettringite
(C6A S 3H32) (Figure 2.10a). Reaction for ettringite formation is as follows:

C3A + 3C S H2 + 26H -----------------> C6A S 3H32

tricalcium gypsum water ettringite
aluminate

Ettringite is the stable phase only while there is an adequate supply of soluble sulfate. A second
reaction takes place if the soluble sulfate is entirely consumed before the C3A has completely
reacted. In this reaction, the ettringite formed initially reacts with the remaining C3A to form a
tetracalcium aluminate monosulfate-12- hydrate known as monosulfate or monosulfoaluminate
(C4A S H12) (Figure 2.10 b).

C6A S 3H32 + 2C3A + 4H -----------------------> C4A S H12

Ettringite tricalcium water monosulfate
aluminate

When the C3A reacts faster than the gypsum can dissolve, the monosulfoaluminate can form
before the ettringite. C2AH8 and C4AH19 phases may still form, even in the presence of gypsum,
though not in sufficient quantities to cause flash set if the C3A is very reactive or there is
insufficient gypsum in the vicinity of the C3A crystal. C4AH19 can form as separate crystals or in
solid solution with C4A S H12.

C3A reaction with gypsum also occurs in two stages with a dormant period similar to that of the
C3S where the reaction rate is distinctly reduced. The underlying reactions, however, are very
different and not obvious. Postulated theories for this reaction include:
1. Buildup of ettringite layer
Ettringite forms by a through-solution reaction and precipitates on the C3A surface
creating a diffusion barrier.82,83
2. Formation of an impervious layer
A water-deficient hexagonal hydrate, stabilized by incorporation of sulfate ions, is
initially formed on the surface of the C3A. This layer then becomes coated with
ettringite.79,84
3. Incongruent dissolution of C3A
C3A dissolves incongruently, leaving an aluminum-rich surface layer onto which
calcium ions are absorbed, reducing the number of active dissolution sites.
Subsequent adsorption of sulfate ions causes a further reduction in the dissolution rate
and slows down the reaction.85
4. Formation of an amorphous gel layer
An amorphous layer is formed on the surface of the C3A that acts as an osmotic
membrane and slows down the reaction.86 The composition of the amorphous layer is
similar to ettringite, and may be a precursor to ettringite formation.87,88

Although the most widely accepted theory is the buildup of an ettringite layer, recent studies
suggest that the formation of an amorphous gel layer slows down the reaction. This amorphous
18
gel is particularly prevalent in the presence of additives. Theories propose that this gel eventually
transforms into ettringite, and that the dormant period ends as a result of a breakdown of the
protective ettringite layer as it converts to monosulfoaluminate. The time required for this
breakdown is dependent on the amount of soluble sulfate from the gypsum being consumed; the
more gypsum in the system, the longer the protective layer remains stable.

C4AF. Hydration of C4AF gives hydration products that are similar to those formed from C3A
under comparable conditions, though typically hydration products of C4AF contain Fe3+ as well
as Al3+. An iron (III) hydroxide gel and calcium ferrite gel are also possible products of C4AF
hydration.89 The reactivity of the pure C4AF is generally much slower than that of the C3A.

2.10.2 Hydration of Cement Phases

Although the basic reaction mechanisms and theories on the hydration of the pure phases pertain
to the phases in cement, there are some significant differences. The initial hydration reactions up
to the time of set are illustrated in Figure 2.11.

Alkalis:
The alkalis, primarily sodium and potassium, are impurities that arise from shales, clays, or the
fuel used in manufacturing cement. Although alkalis are present in small amounts, <1%, they
have a significant effect on hydration. Typically, they exist as sulfates in the form of K2SO4,
Na2SO4, Na2SO4.3K2O (aphthitalite) and/or 2CaSO4.K2SO4 (calcium langbeinite ), and are
usually deposited on the surface of the cement particles. The alkali sulfates dissolve almost
immediately on contact with water. The sulfate ions react with the aluminate and ferrite phases to
form monosulfate and/or ettringite, or they are adsorbed by the C-S-H and an equivalent amount
of OH- are released to maintain the stoichiometric balance with respect to Na+ and K+ ions that
remain in solution. Alkalis can also be present as impurities in the cement phases, with sodium
predominantly found in the aluminate (C3A) phase and potassium more widely distributed in
both calcium and aluminate phases. Typically, these alkalis are released into solution after the
dormant period as the aluminate and silicate phases react (Figure 2.12).

Ca2+ ions are also released into solution during the initial contact of the cement phases with
water. More Ca2+ ions are produced than are required to form the initial surface layer C-S-H or
aluminate phase reactions, and the excess Ca2+ remains in solution, counterbalanced by OH- ions.
The concentration of Ca2+ in solution continues to increase during the dormant period until it
reaches supersaturation, at which point Ca(OH)2 precipitates out of solution. At this point, the
Ca2+ ion concentration in solution is very low. It is the total amount of OH- ions required to
stoichiometrically counterbalance the Na+ + K+ + Ca2+ in solution that gives the cement slurry
its pH 13.90 Typically, the pH is controlled primarily by the Na+ + K+.

Alkalis incorporated into the crystal structure can have a significant impact on the reactivity of
the phases, depending on the quality and quantity of the alkalis. In the case of the aluminate
phase, K+ tends to increase the reactivity, whereas Na+ reduces it.91,92 Reactions of the alite and
ferrite phases are typically accelerated. Alkali sulfate additions generally decrease strength over
the long term.93,94 In the short term, they may either increase94 or decrease strength.93 Alkalis
incorporated in the crystal lattices of the cement phases have no effect on strength
development.92 API Specification 10A for Class G and H cements limits the alkali to 0.75% as
Na2Oe to allow adequate thickening times to be achieved downhole.

19
In cements that are high in K2SO4, reaction between K2SO4 and gypsum in the presence of water
can produce syngenite, KC S 2H. This syngenite can cause lumpiness in the dry cement powder
when stored in humid conditions (> 90% relative humidity) because the KC S 2H binds the dry
cement particles. Precipitation of KC S 2H during cement hydration can cause false or even flash
setting.

Calcium Sulfates:
Gypsum (C S H2) is added to the cement primarily to retard the hydration of the aluminate and
ferrite phases. The effectiveness of the gypsum depends on the rate at which the relevant ionic
species dissolve and come in contact with each other. Thus interground gypsum is far more
effective than interblending the same proportion because intergrinding brings the gypsum
particles into closer contact with the cement particles and produces a shorter diffusion distance
between the two.95,96 Temperature and humidity in the grinding mill can cause the gypsum to
dehydrate, resulting in the formation of hemihydrate (C S H0.5) and/or soluble anhydrite (-C S ).
Hemihydrate or soluble anhydrite can rehydrate to give secondary gypsum, causing a rapid set,
known as false set. This becomes a problem only when the degree of dehydration in the grinding
mill reaches 85 to 90% and depends on the temperature and relative humidity in the mill.
Pumpability can be regained on further mixing or addition of water, assuming the quantity of
secondary gypsum is not too great. Insoluble anhydrite (-C S ) can also be present in the cement
and can form during clinkerization or can exist as a natural component of the gypsum. The rate
of dissolution is: soluble anhydrite > hemihydrate > gypsum > anhydrite; however, the rate at
which the sulfate is available to react with the aluminate and ferrite phases also depends on
whether it rehydrates first to form gypsum. The crystal size also plays a significant role in the
rate of dissolution. Possible dissolution pathways are illustrated in Figure 2.9. Calcium sulfate is
also reported to increase the hydration rate of the alite phase.97-99

Although the primary role of the calcium sulfate is to control the cement setting with respect to
aluminate and ferrite phases, it also affects strength development and volume stability. The
effects on strength are complex and not well understood. Calcium sulfate has an influence on
what calcium alumino hydrates form, their relative percentages particularly at early ages, and the
microstructure at all ages. It is considered that optimum strength values are obtained at an
optimum gypsum value that depends on the clinker quality, cement fineness, hydration time,
and hydration temperature. Alkali sulfates increase the optimum gypsum content.

Cement mineral phases:

As discussed in Section 2.4, the alite (C3S), belite (C2S), aluminate (C3A), and ferrite (C4AF)
phases in cement contain impurities in their crystal structures. Impurities, expressed as oxides,
include Cr2O3, Mn2O3, P2O5, TiO2, B2O3, ZnO, and MgO as well as Na2O and K2O, and can arise
from the raw materials, fluxing agents used during clinkerization, or the fuel. Although the
effects of the impurities on hydration and performance of the cement phases are not well
understood, they can be significant. They include the following:

The alkalis, Na2O and K2O, were discussed earlier in this section. (Section 2.10.2 Formatted: Indent: Left: 0.5", Bulleted +
Level: 1 + Aligned at: 0" + Tab after: 0.25" +
Alkalis). Indent at: 0.25", Tab stops: 0.75", List tab +
Cr2O3 is incorporated preferentially into the alite and ferrite phases, and greatly Not at 0.25"
accelerates reaction at early ages and improves long-term strength.

20
 Mn2O3 tends to increase the size of the alite crystals, making them less reactive and
possibly reducing early strength development.
P2O5 enters the belite structure and can cause a slight lengthening in setting time.
TiO2 is found in the alite and ferrite phases, and retards early strength development
but gives improved strengths with time.
B2O3 forms a solid solution with the belite polymorphs and can retard the setting of
cement if present in sufficient quantity.
ZnO substitutes into the aluminate phase, increasing its reactivity and can cause
gelation problems.
SiO2, although the primary component of the alite and belite phases, is also found as
an impurity in the aluminate and ferrite structures. Its effect on their properties is not
known.
MgO can be present in all phases, occurring preferentially in the ferrite phase,
followed by alite, and in lesser quantities in the belite and aluminate phases. The
MgO does not appear to significantly affect strength or setting properties. In some
cases, the quantity of MgO is greater than can be incorporated into the crystal
structures of the cement phases and it forms as periclase. Periclase can cause
excessive expansion resulting in cracking and loss of strength in set cement because
of its slow reaction rate and expansive properties.
Free lime occurs in clinker as aggregate or single grains, the quantity of which can be
modified by the presence of the impurities. In clinker, the free lime is present as CaO;
however, some of this CaO can be converted during storage or grinding to Ca(OH)2.
Free lime as CaO can reduce strength because of volume instability induced by
conversion to Ca(OH)2 after the cement has set.

The reactivity and performance of cement is a culmination of the effects of the different
impurities, the number of defects in the crystal structure, and the morphology of the crystal
structure of the different phases. This is why cement can vary not only from one source to
another but also among batches from the same source. Typically, the percentage of impurities in
cement is very low and as their resultant effects are taken as inherent to cement performance,
routine analysis on these minor or trace elements is not performed. However, alkali, MgO, and
free CaO can have a significant and deleterious effect on cement performance, and contents are
determined according to ASTM C 114(24) and API(13) specifications.

Ferrite phase:
In cement, the ferrite phase is typically a solid solution having composition Ca2(AlxFe1-x)2O5
where 0<x<0.7. The reactivity of the ferrite phase can vary over a wide range and appears to be
dependent on the Al/Fe ratio where, generally, the lower the Fe content, the higher the
reactivity.100 The ferrite phase typically contains about 10% impurity and its composition
approximates to Ca2AlFe0.6Mg0.2Si0.15Ti0.05O5.101 This composition is lower in Fe than in pure
C4AF, and is consistent with its higher reactivity. In the presence of certain additives, the ferrite
phase can be as reactive as the C3A phase.

Hydration products:
The major product of cement hydration is an amorphous C-S-H gel of variable C/S and H/S
ratios, as is the case with pure C3S and -C2S. However, the C-S-H gel in cement may also
incorporate a limited amount of impurities such as Mg2+, Al3+, K+, Na+, and SO42. The
morphology of the C-S-H also differs, as can be seen in Figures 2.6 and 2.7. Hydration of pure
C3S shows thin films of C-S-H both around the particles and filling the pore space, whereas in
21
the cement, the C-S-H forms a thick gel layer around the cement particles.102 Hexagonal plates of
Ca(OH)2 are also observed in the hydrated pure C3S, but not in the cement, at three hours. The
morphology of the hydration products differs even more at seven days, when the cement shows
not only the silicate hydration products but also those of the aluminate and ferrite.

The hydration products from aluminate and ferrite phases also contain a significant amount of
impurities. The Al3+/Fe3+ ratio in the hydrated phases also varies, depending on the amount of
aluminate and ferrite in the original cement and the degree of hydration. Typically, ettringite and
monosulfate are denoted as AFt (trisulfate) and AFm (monosulfate) phases because of the
variable Al3+/Fe3+ ratio and because of the impurities in the structures. The AFt formed in
hydrated cement tends to be relatively well crystallized, whereas the AFm tends to be poorly
crystallized.

Minor products found in hydrated cement may include a hydrotalcite-like

[Mg3Al(OH)8](CO3)0.5.2H2O phase that is formed from the Mg2+ impurities found in cement
phases,103, 104 and brucite [Mg(OH)2], which may also be present if periclase exists in the clinker.
A poorly crystalline siliceous phase containing a significant fraction of iron has been found and
is thought to be a precursor to hydrogarnet.

Composition of the aqueous phase:

The composition of the pore solution extracted from set cements indicates that the solution
contains relatively high concentrations of Ca2+, K+, Na+, SO42-, and OH- almost immediately on
mixing cement with water.(105-108) This concentration remains relatively constant during the
dormant period and the end is marked by a rapid drop in the Ca2+ and SO42- and an increase in
the Na+ and K+ (Figure 2.12). This occurs within 6 to 16 hours, depending on the cement and
water-to-cement ratio, and is associated with renewed hydration of the cement phases. The
solution that ultimately is in equilibrium with the set cement is that of the alkali hydroxides.
Concentrations of Al3+ and silicate species remain low throughout.88,107,108 Solution compositions
can be modified with additives.88,108

Cement phase hydration interactions:

Cement is a composite material in which alite is the predominant phase, and belite is bound in a
matrix composed of ferrite and aluminate phases. The ferrite phase is closely mixed with the
aluminate, and orientated intergrowths between the two can occur. Furthermore, Fe3+ and Al3+
are incorporated in small amounts in the alite and belite, and Si4+ is found in the aluminate and
ferrite phases. On hydration, the dissolution of ions and the proximity of the phases is such that
the reactions between phases are not independent. This is evident in Figure 2.6 in that the
particle surfaces are covered completely by gel. There is no evidence of Aft (ettringite) needles
as shown in the hydration of pure C3A and C S H2 (Figure 2.10). In the early ages of hydration,
the C-S-H gel in the cement engulfs and perhaps nucleates on the AFt needles and completely
covers the cement particles.102 At high water content, a gel layer with composition similar to that
of precursor AFt gel forms over the grain surface.109

2.10.3 Hydration Reactions Above 104F (40C)

Previous sections relate primarily to the hydration mechanism that occurs during initial mixing,
and is the same whether the temperatures are sub-zero or 104F (40C). The rate of hydration of
the cement phases, however, increase with increasing temperature, and the resulting thickening
and setting times consequently decrease. The effect of temperature on the aluminate phase
reactions is not clear and appears contradictory. The consensus is that the chemical factors
22
controlling the balance between the AFt and AFm phases are influenced by changes in the
sorption of the C-S-H to sulfate ions.90,110,111 The AFt and AFm phases become unstable, and
eventually revert to an amorphous gel product at temperatures of 122 to 212F (50 to 100C).
There is no indication that the hydrogarnet phase forms at these temperatures. The
decomposition of the AFt and AFm phases does not appear to affect the setting properties,
compressive strength, or volume stability of cement, although it could contribute to gelation or
setting problems with certain additives.

The amorphous C-S-H gel, the primary binder in cement, undergoes significant, albeit gradual,
changes as the temperature increases. At about 122F (50C), the C-S-H becomes more lime-rich
and there is less water bound in the structure. Effectively, the Ca/Si ratio of the C-S-H increases,
whereas the H/S ratio decreases.110,112 The degree of polymerization of the silicate ions increases
and the C-S-H gel starts to become denser with a coarser microstructure.112,113 The resultant
effect of the decrease in bound water and coarser microstructure is an increase in porosity, and a
decrease in ultimate compressive strengths. C-S-H may also incorporate more of the Al3+ and
SO42- as temperatures increase, accelerating the AFt and AFm decomposition.

Above 230F (110C), the hydration products formed differ considerably from those obtained at
lower temperatures. Alite and belite phases hydrate to give crystalline -C2SH rather than
amorphous C-S-H, while the aluminate and ferrite phases produce a hydrogarnet phase or
strtlingite. -C2SH is a relatively dense crystalline material that is porous and weak and is
deleterious in that it provides high permeability and low compressive strength. Its formation has
long been associated with strength retrogression. Formation of -C2SH can be prevented, or at
least minimized, by adding finely ground silica, such as silica flour, to the cement. The silica
reacts as a pozzolan, reducing the calcium hydroxide content and resulting in formation of
calcium silicate hydrates that range in crystallinity from almost amorphous to highly crystalline,
depending on temperature, pressure, amount and type of silica added, and reaction time. The
phase equilibrium of the calcium silicate hydrate phases that form between 248 and 752F (120
to 400C) have been well studied for the pure CaO-SiO2-H2O system.114-124 Typical calcium
silicate hydrates that have been identified under hydrothermal conditions are listed in Table
2.12, and an example of the phase equilibrium based on temperature and total C/S ratio is
provided in Figure 2.13. In principal, the phases do not form directly but through a series of non-
equilibrium reactions that usually start with the poorly crystalline C-S-H phases and various
crystalline phases that are usually formed at the lower temperatures. The phases formed depend
primarily on the C/S ratio. Many factors can modify the reaction sequence and the stability limits
of the phases formed, however. These factors include:

Pressure115,117The upper stability temperature for a given phase can be increased by Formatted: Indent: Left: 0.5", Bulleted +
Level: 1 + Aligned at: 0" + Tab after: 0.25" +
212F (100C) under increased pressure. Shifts in the temperature-pressure curves Indent at: 0.25", Tab stops: 0.75", List tab +
can also result in crossover points in the upper stability of the phases with Not at 0.25"
corresponding changes in the reaction sequence.

23
 Temperature gradients124Silica migration from hotter to cooler zones by liquid or Formatted: Indent: Left: 0.5", Bulleted +
Level: 1 + Aligned at: 0" + Tab after: 0.25" +
vapor transport can modify the C/S ratio and hence phases formed. Indent at: 0.25", Tab stops: 0.75", List tab +
Solubility of the silica phase119,124-126 A more soluble, reactive silica (microsilica) Not at 0.25"
source initially produces lower C/S ratio phases and gives a metastable equilibrium.
These phases may transform with time to the stable phases associated with the bulk
C/S ratio. A less reactive crystalline form (silica flour) gives a more true
equilibrium and tends to follow the phase sequence for the bulk C/S ratio. Reactive
silica sources generally give lower compressive strengths and also show strength
retrogression at higher temperatures.
Persistent metastable phase124This phase is metastable with respect to the C/S
ratio and does not convert to the stable phase under typical exposure times.
Impurities115,127-129These substitute for ions in the crystalline phases and can
increase the upper stability limits.

Normally in oilwell cementing, approximately 35% silica in the form of silica flour is used to
prevent strength retrogression that can occur at temperatures above ~ 248F (120C). This
percentage of added silica gives an effective C/S ratio in the cement blend of approximately 1.0.
Initially, C-S-H(I) forms and is usually intermixed with a poorly crystalline tobermorite. As
temperatures increase, the tobermorite becomes more crystalline and at around 302F (150C), it
becomes the predominant phase (Figure 2.14). In terms of oilwell cementing, this phase is
typically associated with good compressive strength and low permeability. Above 200C, the
tobermorite becomes unstable and transforms into xonotlite or a mixture of xonotlite plus
gyrolite. Gyrolite converts to truscottite at temperatures above 482F (250C). Both xonotlite and
truscottite convert to wollastonite or pseudowollastonite at temperatures of 725F (385C). These
phases are all associated with relatively good compressive strengths and low permeability,
though not as good as tobermorite. Generally, over time, the permeability will increase slightly
and the compressive strength will decrease as the phases increase in crystallinity, with the
crystals becoming larger and more defined. There appears to be an optimum degree of
crystallinity for maximum performance.130

The crystalline calcium silicate hydrate phases can incorporate the impurities found in cement as
well as Al3+, Fe3+, and SO42- to a degree, without altering the structure significantly. Given a
sufficiently high concentration of ions, the calcium silicate hydrate phases convert to a different
phase or mixture of phases. These phases can be similar in structure to the original phase or
completely different. Depending on the amount of substitution or phases formed, the effects on
performance can range from highly beneficial to deleterious. Tobermorite, for example, with 7%
alumina substitution, produces an optimum compressive strength, whereas alumina substitution
above 8% results in the formation of other alumina-containing phases such as CAS2 (anorthite),
C2ASH (bicchulite), CAS4Hx (wairakite), AH (boehmite), and Si-rich hydrogarnets in increasing
amounts and the strength subsequently decreases.128,131 Scawtite115,131-133 is a calcium silicate
carbonate hydrate that forms readily in trace amounts under hydrothermal conditions from
contact with aqueous solutions containing dissolved CO2. It is also formed when sodium
carbonate or finely ground calcium carbonate is added at levels of 10% or less to a 35% silica
cement slurry. The crystal structure of scawtite is similar to that of xonotlite in which 5.3% CO2
replaces part of the silica. Scawtite in trace amounts enhances cement performance, but when
present in significant quantities, it produces brittleness and poor performance. Pectolite127,134 is a
sodium-substituted calcium silicate hydrate and forms as replacement to both tobermorite and
xonotlite when they are exposed to salt solutions like those found in high-temperature brines. It
24
has a somewhat higher permeability and lower compressive strength than tobermorite, and
typically performs like xonotlite. Reyerite135, which is occasionally found in well cements, is an
alumina and alkali-substituted truscottite with similar performance properties.

The potential for using substituted-calcium silicate hydrates as stable binders has also been
reported.133,136 Of particular interest are the lime-magnesia-silica-water and lime-alumina-silica-
water. In the magnesia system, the compositions at or near the diopside stability field showed the
highest compressive strengths at 300 to 400C. Xonotlite and diopside both formed, though with
time, the xonotlite content decreased and diopside increased. Alkali, up to 10 weight percent, had
little effect on the phases formed. The anorthite region in the lime-alumina-silica-water system
produced two polymorphs of anorthitea modified hexagonal form and a triclinic formthe
relative amounts of which depended on the starting materials. Hexagonal anorthite produces
favorable cementitious properties over the 572 to 752F (300 to 400C) temperature range,
whereas the triclinic form is detrimental in that it has no binding property. The non-cementitious
triclinic form is also favored where alkali salts are present. It is not clear whether the modified
hexagonal or triclinic anorthite is the stable form at these temperatures. If the triclinic form is the
stable form, strength retrogression would result over the long term.

Fly ash has often been considered a potential source of silica for hydrothermal systems. It is, in
fact, an aluminosilicate glass and, in effect, falls into the lime-alumina-silica water system.
Because the alumina content is relatively low, it is usually incorporated in the structure of the
calcium silicate hydrates, such as the tobermorite or xonotlite phases. The alumina-to-silica ratio
of fly ashes from different sources varies considerably, as does the reactivity of the
aluminosilicate glass,137 and this clearly impacts the phases formed and their stability fields. As a
result, the fly ash, if used as a source of silica, can give properties that range from good to
deleterious.

Typically, cement compositions are formulated so that tobermorite is formed as the stable binder
at temperatures up to 572F (300C), with xonotlite and possibly gyrolite or truscottite forming at
higher temperatures. However, in actual wells, the standard conditions for equilibrium
transformations are more or less nonexistent and thus the phases formed will depend on the
actual characteristics of the well. The phases most likely include the slightly modified calcium
silicate hydrates, the substituted-calcium silicate hydrates, as given in Table 2.12, plus various
aluminosilicate and calcium aluminate compositions.

2.10.4 Corrosive Environments

Set cement by its very nature contains a degree of porosity and permeability. As such it is
susceptible to chemical attack by formation waters. This can lead to loss in cement integrity in
the long term. In HTHP wells corrosive effects from formation waters can have a greater impact
on long term cement integrity than other reservoir conditions.

Sulfate (SO42-) attack:

Oilwell cements used in most cementing operations are sulfate-resistant and conform to the MSR
and HSR classification as specified in API Specification 10A. Formation waters containing
sulfates, such as sodium or magnesium sulfate, can attack the set cement and cause expansion,
cracking, strength loss, and, in extreme cases, disintegration.138-139 Sulfate attack is caused by the
reaction of the sulfates in the ground waters with the calcium hydroxide (Section 2.10.1) in the

25
cement to form gypsum, which in turn reacts with the AFm (monosulfoaluminate) phase in
cement to form AFt (ettringite), according to the following equations (See also Figure 2.9):

Ca2+ + SO42- + 2H2O C S H2

C4A S H12 + 2C S H2 + 16H C6A S 3H32

The volume occupied by the AFt, and to a lesser extent gypsum, is larger than that of the original
reactant, which causes disruptive expansion. The gypsum can also react with any unreacted
aluminate or ferrite phase to produce Aft, though the main source is likely to be the AFm phase.
In more severe cases, calcium required for the attack to proceed comes from dissolution of the C-
S-H. MgSO4 attack is more aggressive than either Na2SO4 or K2SO4 as the Mg2+ reacts with the
OH- produced from dissolution of the CH or decalcification of the C-S-H to form brucite:

Mg2+ + 2OH- Mg(OH)2

This reaction decreases the amount of OH- in solution, but further dissolution of CH or
decalcification of the C-S-H replenishes the OH-. Formation of brucite also causes expansion.
Sulfate attack is normally only a problem where ettringite is present and bottomhole static
temperatures (BHST) are below 140F (60C).

Resistance to sulfate attack is increased when the cement powder is modified by replacing the
aluminate with ferrite. This replacement reduces both the amount of ettringite formed on
hydration and the amount of free lime. Addition of pozzolanic materials, such as fly ash, also
reduces sulfate attack because the materials react with the CH in cement and render it
unavailable for reaction.

Carbonate (CO2) Attack:

Carbon dioxide corrosion in cements is a well known phenomena that typically results in
increased porosity and permeability that adversely affects the cements structural integrity and in
the case of oil well cements cam cause lost zonal isolation.140-142 Carbonate attack occurs when
CO2 dissolves into the pore solution of set ordinary Portland cement and reacts with Ca2+ ions to
produce insoluble CaCO3. The basic reactions of which may be represented as follows:

CO2 + 2OH- CO32- + H2O

Ca(OH)2 + CO32- CaCO3 + 2OH-
C-S-H gel + CO32- CaCO3 + hydrated amorphous silica

The OH- and Ca2+ initially come from dissolution of the CH. When all the CH has been
consumed further OH- and Ca2+ are obtained on decomposition of the hydrated silicate and
aluminate phases. Decomposition of C-S-H results in a decalcification, initially by lowering the
Ca/Si ratio and ultimately yielding a highly porous form of silica gel that can change with time to
crystalline silica.140,141 The monosulfate and ettringite decomposition results in the formation of
alumina and iron (III) oxide gels with the alumina gel eventually crystallizing to gibbsite
(Al(OH)3). Although the calcite form of CaCO3 is the most commonly encountered product,
other forms of calcium carbonate may form such as aragonite and vaterite. The reaction also
results in a decrease in the aqueous pore fluid pH with values reaching 8.5 or less.143 The overall
reaction results in a deterioration of the cement structure and this can cause serious damage to
down hole tubulars and destroy zonal isolation integrity.

26
As more CO2 invades the cement matrix several other reactions take place.141,142 Calcium
carbonate formed during the initial reactions can react with the CO2 to form the much more
soluble calcium bicarbonate [Ca(HCO3)2], which can migrate out of the cement matrix.
Dissolved Ca(HCO3)2 can further react with calcium hydroxide, forming additional calcium
carbonate and fresh water. The liberated water can then dissolve more calcium bicarbonate
according to the following reactions

CO2 + H2O + CaCO3 Ca(HCO3)2

Ca(HCO3)2 + Ca(OH)2 2CaCO3 + H2O

These reactions perpetuate as long as there is CO2 available and will continue to destroy the
cement matrix. Furthermore the greater the concentration of CO2 the greater the rate of
carbonation.144 High temperatures associated with HTHP wells also accelerate these reactions
causing even faster corrosion and lost cement integrity.142

Resistance to CO2 attack can be improved by reducing cement permeability or reducing the
amount of cement exposed. Cement permeability can be reduced by reducing the water content
or adding pozzolanic material such as fly ash, natural pozzolans or silica fume. The pozzolanic
materials also decrease the amount of CH available to react with the carbonate ion. In extreme
cases, depending on economic viability, non-Portland cement systems such as epoxy resins or
sodium-polyphosphate-modified fly ash/calcium aluminate blended cements146 can be used.
These are not susceptible to CO2 attack..145,146

Hydrogen sulfide (H2S) attack

Formation waters that are saturated with H2S (sour gas) or any natural gas containing HsS at
temperatures > 350F causes cement to deteriorate with loss in cement integrity. H2S is oxidized
to sulfuric acid which then dissolves the cement leaving the cement matrix weak and brittle.147
The reactions representing H2S and cement phases are represented as follows:

H2S + Oxidation H2SO4

H2SO4 2H+ + SO42-
2H+ + SO42- + Ca(OH)2 CaSO4 +2H2O
6H+ + 3SO42- + 3 C-S-H 3CaSO4 + 2(SiO2.nH2O) + 6H2O

Methods used to improve resistance to H2S attack are similar to those used in prevention or
minimizing CO2 attack.146,148

2.11 Additives
Because of the inherent nature of Portland cements and the demands placed on them during the
cementing process, the performance properties must be modified to meet the demands of a
particular well application. Chemical compounds, commonly referred to as additives, are used to
alter the hydration chemistry in terms of:
the initial C-S-H gel Formatted: Indent: Left: 1.5", Bulleted +
Level: 1 + Aligned at: 0" + Tab after: 0.25" +
the aluminate or ferrite reactions Indent at: 0.25", Tab stops: 1.75", List tab +
growth of CH Not at 0.25"

the surface charge

ionic composition of solution

27
The effect of the additives on the C-S-H gel can be either by substitution of calcium or silicon in
the C-S-H gel structure, or through absorption into the C-S-H gel, or by adsorption onto the C-S-
H gel. The intended result is to either decrease or increase the permeability of the C-S-H gel.

Aluminate or ferrite reactions (Figure 2.9) are commonly modified by

incorporating a competing sulfate species that can prevent the growth of the AFm or AFt Formatted: Indent: Left: 0.08", Bulleted +
Level: 1 + Aligned at: 0" + Tab after: 0.25" +
phases Indent at: 0.25"
sequestering or chelating the aluminum or iron to prevent the formation of AFm or AFt
modifying the precursor gel by absorption to either increase or decrease the permeability
change the solubility limits through changing the solution chemistry

The growth of the CH may be inhibited by adsorption of poisonous additives onto the crystal
structure.

The surface charge distribution of the initially formed hydration products (C-S-H gel, precursor
gel, AFm or AFt phases) can be altered by absorption of suitably charged additives (Figure
2.15). The average molecular weight (Mw) of the charged additives also plays a significant role.
For example, a polymer can act as a surfactant, a retarder, a dispersant, or a viscosifier or fluid-
loss control agent, depending on its molecular weight (surfactant behavior is associated with a
very low molecular weight; viscosification is associated with a very high molecular weight).

Ionic composition of the solution also plays a significant role in the hydration process (Figure
2.11). Additives that alter the ionic composition significantly alter the hydration process. For
example, adding soluble sodium or potassium salts will precipitate CH from solution and reduce
the dormant period.

Just as the hydration reactions in cement are not independent of each other, neither are the
reactions associated with additives. Usually, one or more of the effects listed above will occur
within a slurry with a given additive.

There are a number of chemical compounds that have been proven effective in modifying the
properties of Portland cement slurries. Used alone, these compounds have a primary effect on the
cement slurry that is considered to be beneficial. The same compounds will also exhibit at least
one secondary characteristic that may be either beneficial or detrimental to slurry performance.
Typically, detrimental effects are reduced or beneficial effects are enhanced through
modification of the additive or use of additional additives. For most downhole requirements,
more than one additive is needed. This give-and-take relationship between additives is the basis
of cement slurry design.

The reaction of these additives with the cement and the interaction between additives is not well
defined from a chemistry point of view. Several theories exist and many studies have been
performed with conflicting interpretations. What is actually known are the physical effects of
these additives on slurry performance properties. The slurry performance properties that are
measured include:

Thickening time. The time required for a cement slurry of a given composition to reach a
consistency of 100 Bearden units of consistency (Bc).
Compressive strength. The degree of resistance of a material to force acting along one of the
axes in a manner tending to crush it, usually expressed as force per unit area of surface.
28
 Rheology deals with the measurement of deformation and flow of matter. The important
parameters are the plastic viscosity and the yield point. Plastic viscosity is measured in
centipoise; yield point is measured in force per unit of contact area.
Fluid loss is the volume of filtrate lost to the permeable medium through filtration. In
cementing applications, this value is the volume of filtrate lost from the cement slurry at a
given temperature over a period of 30 minutes. The measured value is reported in units of
cubic centimeters per 30 minutes.
Free fluid is the volume of supernatent fluid developed at the vertical surface of a column of
fluid (i.e., cement slurry) as a result of the settling of suspended solids to their minimum
stacking volume. The unit of measure is reported in milliliters or as a percentage of column
height.
Slurry stability relates to the uniformity of consistency of a slurry. Several methods exist for
determining the stability of a slurry.
The sedimentation test is used to construct the density profile of a sample that has been Formatted: Indent: Left: 0.25", Bulleted +
Level: 1 + Aligned at: 0" + Tab after: 0.25" +
allowed to set under a given temperature and pressure. The density of the sample Indent at: 0.25", Tab stops: 0.5", List tab +
segments from top to bottom are determined according to the Principle of Archimedes Not at 0.25"
and values are quoted as percentage density difference.
Solids suspension determines the stability of the cement slurry during and following
cement slurry placement. At predetermined times during a modified thickening time test,
stirring is ceased, for a period of 10 minutes. The maximum deflection in Bearden units
of consistency, at the instant of start up, gives and indication of momentary solids
settling. Evaluation of the consistency of the slurry after testing also gives an indication
on settling and stability.
Dynamic settling test uses a modified high temperature, high pressure consistometer and
paddle. The slurry is stirred at a reduced speed of 25 rpm for 30 minutes after reaching
the desired bottomhole circulating temperature conditions according to API standards.
Stability is determined on the consistency of the fluid portion of the slurry poured into a
beaker and on the cone shaped build-upof solids settled on the bottom plate of the
modified paddle.
Free fluid test can also give an indication of slurry stability based on the amount of
particle settling in the slurry column.

Methods for measuring these properties are outlined in the API Recommended Practice for
Testing Well Cements 10B.25 The additives used to modify the properties of cement slurries for
oilfield well-cementing applications fall into the following broad categories.

accelerators Formatted: Indent: Left: 0.5", Bulleted +

Level: 1 + Aligned at: 0" + Tab after: 0.25" +
retarders Indent at: 0.25", Tab stops: Not at 0.25"
extenders
weighting agents
dispersants
fluid-loss control agents
lost-circulation agents
strength retrogression prevention agents
free water/free fluid control
expansion agents
special additives

29
The demand for new additives with special properties and tuned performance continues to
increase. These demands include, but are not limited to, density range of application, temperature
stability, economics, viscosity range, singular function, multifunction, rate of solubility,
synergism with co-additives, and resistance to cement variability.

2.11.1 Accelerators
The term accelerator indicates a speeding up or shortening of the reaction time required for a
cement slurry to become a hardened mass. This indicates a reduction in thickening time and/or
an increase in the rate of compressive-strength development of the slurry. This acceleration is
accomplished by adding compounds that increase the reactivity of aluminate (substituted -C3A)
or alite (substituted C3S) phases or both and, in most cases, by using an additive that changes the
aqueous phase composition (i.e., pore solution) of the cement slurry. Acceleration is particularly
beneficial in cases where a low-density (i.e., high water content) cement slurry is required or in
low-temperature formations.

The compounds found to be most effective are composed of a cation or anion, or both, that is
similar to the components of the cement (Table 2.13). The main exception is the chloride anion,
which is not a natural component of Portland cement but one that will substitute for the sulfate in
the aluminate hydration phases.

Calcium Chloride (CaCl2 ). Of the chloride salts, CaCl2 is the most widely used and, in most
applications, the most economical. The exception is when water-soluble polymers such as fluid-
loss control agents are used (Section 2.11.6). The major benefits of CaCl2 are the significant
reduction in thickening time achieved and the fact that, regardless of concentration, it always acts
as an accelerator.

CaCl2 is available in both solid and liquid forms. The solid form of CaCl2 is white and is
available in flake, powder, and bead forms. The liquid form is a clear aqueous solution with a
weight average activity that ranges from 37 to 40%. The flake form is typically a technical grade
material with an average activity of 77 to 80 wt %. The powder form has an activity of about 96
wt %. The bead form is typically anhydrous, and has an activity of 94 to 97 wt %. Because of the
hygroscopic nature of the anhydrous bead form, the flake or the powder form is normally used.
The effectiveness of the flake, powder, or bead form is controlled by its activity. Direct
substitution of one form for another without proper testing is not recommended. This is
particularly important with slurries for which the thickening time has been intentionally
minimized.

CaCl2 is typically used at concentrations of 1 to 4% by weight of cement (BWOC). Above a

concentration of 6% BWOC, the results are unpredictable and gelation can occur. Another
concern is CaCl2s limited sensitivity to increased concentrations. Often, little improvement in
properties is obtained by increasing the concentration. In freshwater slurries or seawater slurries
of average density, CaCl2 concentrations of 2% BWOC or less are recommended (Figure 2.16).
In extended slurries, CaCl2 concentrations of greater than 2% BWOC are typically required due
to the dilution caused by the high water requirements of extended slurries. Furthermore, at low
temperatures (5C), performance of CaCl2 is concentration-dependent and, at higher
concentrations, its performance becomes unpredictable149.

Sodium Chloride (NaCl ). NaCl, or common table salt, is the second most widely used and
most versatile of the chloride salts. Depending on its concentration, NaCl can act as an

30
accelerator or a retarder, and it acts a mild dispersant at all concentrations. Some additional uses
for NaCl are to
improve bonding to pipe
stabilize reactive formations (i.e., shale and gumbo)
enhance bonding to salt formations
reduce the permeability of set cement
improve the durability of set cement in contact with saltwater-containing formations
increase slurry density without the use of dispersants or reduction in water content

In general, NaCl acts as an accelerator at concentrations from 1 to 10 % BWOW, although the

most commonly used concentration of NaCl as an accelerator is 3% BWOW. At concentrations
of 10 to 15% BWOW, the accelerating effect is minor, and at 15 to 18% BWOW, little effect is
observed. At a concentration of 20% BWOW or higher, NaCl acts as a retarder (Figure 2.17).
Across the entire concentration range, the one common effect is a decrease in slurry rheology.

Potassium Chloride (KCl ). KCl is similar to NaCl as an accelerator. KCl has two advantages
over other accelerators: 1) its stabilizing effect on shale or active clay-containing formations, and
2) its minimum effect on the performance of fluid-loss additives. KCl may be used as an
accelerator at concentrations of up to 5% BWOW and for formation stabilization at 3% BWOW.

KCl is commercially available in solid or liquid form. It will accelerate the thickening time and
the rate of compressive-strength development of cement slurries. KCl also provides a mild
dispersing effect. Above concentrations of 5% BWOW, the results will become unpredictable
and slurry gelation can occur.

The effectiveness of chloride salts as accelerators is due to the dual effect of alkali (Section
2.10.2) and chloride on the hydration of cement phases. The mole fractions of the chloride anion
in NaCl, KCl, and CaCl2 are 0.5, 0.5, and 0.67, respectively. The chloride ion accelerates the
aluminate and ferrite phases after the gypsum has reacted.150 The chloride ion reacts to form the
hydrochloroaluminate (3CaO.Al2O3.CaCl2.xH2O) and stabilizes the AFt. The calcium ion
primarily affects the reactivity of the alite phase and, to a lesser extent, the aluminate and ferrite
phases. The acceleration on the alite phase is essentially catalytic because of the increased
availability of calcium in solution to initiate the precipitation of CH (Figure 2.11). The sodium
and potassium do not tend to affect the alite hydration. Sodium reduces the reactivity of the
aluminate phase; potassium increases it. 91,92 Approximately the same reduction in thickening
time can be achieved with the addition of either 1% CaCl2 BWOC or 5% NaCl BWOW (Figure
2.18). However, the early compressive strength of the NaCl slurry will be slightly lower. A
further benefit of NaCl, as compared with CaCl2, is the increased solubility of polymers in NaCl-
containing solutions.

Sodium Silicate (Na2SiO3). Sodium silicate is normally considered to be a chemical extender,

although it also functions as an accelerator. The acceleration effect is caused by the precipitation
of calcium from solution to form C-S-H. The effectiveness is dependent upon the concentration
and molecular weight. The low molecular weight form may be used at concentrations of 1%
BWOC or less to accelerate normal-density slurries. The high-molecular-weight form is an
effective accelerator at concentration of up to 4% BWOC. The advantages of the high-molecular-
weight form are the negligible effect on fluid-loss control additives and a rapid set.151

31
Seawater is a naturally occurring source of alkali chloride salts. Alkali chloride salts include
those mentioned above plus magnesium chloride. The composition of seawater varies widely
around the world (Table 2.14). For example, the equivalent chloride salt content can vary from
2.7% to 3.8% BWOW.152

Alkali Hydroxides (Ca(OH)2, NaOH). Alkali hydroxides are commonly used in pozzolan-
extended cements. They accelerate both the pozzolanic material and the cement component by
altering the aqueous chemistry. Alkali hydroxides accelerate the CH supersaturation, resulting in
a decreased dormant period according to the nucleation theory (Sec. 2.10.1). The accompanying
increase in pH results in the dissolution of the pozzolanic material153 and accelerates the
hydration product reaction, Section 2.10.2. Hydrated lime (CaO), containing water of
crystallization, is also used as an accelerator in pozzolan-extended cements and functions in the
same manner as the alkali hydroxides.

Mono-Calcium Aluminate (CA). Calcium aluminate is used as an accelerator in pozzolan- and

gypsum-extended cements. The CA increases the aqueous phase concentration of calcium and
aluminum, and reacts with the gypsum, thereby increasing the aluminate/ferrite-to-gypsum ratio.
This increases the aluminate and ferrite reaction and, depending on the amount of gypsum
consumed, produces the reaction products as illustrated in Figure 2.9. Concentration is limited to
5% BWOC or less. At higher concentrations, there is insufficient gypsum to prevent flash set.

2.11.2 Retarders
The cements commonly used in well applicationsAPI Class A, C, G, and H as produced in
accordance with API Specification 10A(13), do not have a sufficiently long fluid life (thickening
time) for use at temperatures above 100F (38C) BHCT. For extending the thickening time,
additives known as retarders are required.

The chemical mechanisms of cement retardation are complex and not well understood. Whereas
acceleration is accomplished by modifying the aqueous phase composition, retardation is usually
accomplished by modifying the hydration products. On the basis of the theories proposed
concerning the mechanisms of the dormant period (See Section 2.10.1), an additive that would
make the protective layer around the cement grain less permeable and/or poison the growth of
the CH nuclei would increase the dormant period. Different mechanisms have been proposed for
modifying the protective layer and include adsorption of retarder onto the protective layer,
precipitation of retarder with Ca2+ and/or OH- to form an impermeable layer, and adsorption of
the retarder onto the nuclei of the hydration products, preventing further growth and
transformation into the second hydrate. All or some of these mechanisms may apply in most
cases.

Lignosulfonates. Of the chemical compounds identified as retarders, lignosulfonates are the

most widely used. A lignosulfonate is a metallic sulfonate salt derived from the lignin recovered
from processing wood waste. Lignin is a phenylpropane (Figure 2.19) polymer of amorphous
structure comprising 17 to 30% wood and is closely associated with the holocellulose that makes
up the balance of woody material. Holocellulose can be separated from the lignin fraction only
by chemical reaction. The reaction route used is the Kraft pulping process in which sodium
sulfite reacts with the lignin under pressure at 170C to separate the holocellulose. The resultant
aqueous solution of lignosulfonic acid, simple sugars, starches, and natural gums is known as
lignin liquor. The composition of the liquor is dependent on the wood source and the reaction
32
conditions. Performance of the liquor as a retarder is dependent upon the proportioning of these
compounds, their molecular weights, and the degree of sulfonation. Lignosulfonates have
average molecular weights of 20,000 to 50,000, and exist in solution as polyelectrolyte spherical
microgels with charges that exist predominantly on the outside of the sphere.

Calcium and sodium lignosulfonates, the most common lignosulfonates, are available in solid or
liquid form and range in color from light brown to very dark brown. The salt and the
composition of the lignosulfonate liquor have a pronounced effect on the performance properties
of the compound. Raw liquors are generally not used in well applications because when the
liquor is incorporated into a cement slurry, the natural gums and starches that exist as suspended
partially solubilized particles can result in a progressive and undesirable viscosification. As a
result, the lignosulfonate liquors commonly used in well cementing applications are filtered to
remove the swollen or insoluble components.

As a result of processing, three grades of lignosulfonate are available for the retardation of
cement slurries. Each grade is available as calcium/sodium or sodium salts. The three grades can
be described as filtered, purified, and modified.

The filtered grade is that described previously. The calcium or sodium salt is typically used at a
temperature of 200F BHCT or less at a concentration of 0.6% BWOC or less. It may be used at
higher temperatures if economics permit. A disadvantage is that the performance of filtered-
grade lignosulfonate is more dependent on cement quality than the purified or modified grades.
In some cases, gelation may occur. However, when filtered-grade lignosulfonate is used with
cements of adequate quality, the performance of fluid-loss additives is generally enhanced as a
result of the sugar content.

The purified grade represents a class of lignosulfonates in which the sugar content has been
reduced. The calcium/sodium salt is typically used at a BHCT of 200F or lower and at a
concentration of 0.5% BWOC or less. The sodium salt is generally effective from 200F to 300
F and at a concentration of 0.1 to 1.0% BWOC. The response is generally more predictable than
that of the filtered grade, and the dispersing effect is more pronounced. A disadvantage is that the
performance of fluid-loss additives is generally not improved.

The modified grade represents lignosulfonates that have been blended or made to react with a
second component. The compounds most commonly used as blend components are boric acid
and the hydroxycarboxylic acids or their salts. Blended materials are available as calcium or
sodium salts. The modified lignosulfonates are typically used at a BHCT of 200F or above.
They are more effective than the purified grade at temperatures greater than 250F. The
advantages are their improved high-temperature stability above 300F BHCT, increased
dispersing activity, and synergism with fluid-loss additives. The increased dispersing activity of
modified lignosulfonates, however, can be a disadvantage.

Numerous studies on the mechanism of retardation using lignosulfonates on both pure phases
and cement have been reported.154-158 The complexities of the hydration, coupled with the
different compositions in lignosulfonate, whether filtered, purified or modified, give rise to
different and often conflicting opinions. Considering only the purified lignosulfonate, one can
reasonably assume the following:

33
 Hydration of both aluminate and alite phases are retarded, with the effect being more Formatted: Indent: Left: 0.5", Bulleted +
Level: 1 + Aligned at: 0" + Tab after: 0.25" +
predominant on the alite phase. Retardation increases with the increase of the Indent at: 0.25", Tab stops: 0.75", List tab +
alite/aluminate ratio.154-157 Not at 0.25"
Lignosulfonate modifies the morphology of the AFt needles, with smaller needles
forming. Conversion of AFt to AFm is considerably retarded.159
C-S-H morphology is modified, and the crystal growth of CH is inhibited.160
Both the nucleation and adsorption theories explain the effects of the lignosulfonate. It is
considered that the sulfonate and hydroxyl groups of the lignosulfonate adsorb onto the C-S-H
gel and CH, preventing further growth. The sulfonate groups probably also replace sulfate in the
AFt phase, which would prevent the needles from growing and would stabilize against
conversion to AFm. The slower rate of formation of AFt and the decrease in the size that
prevents the interlocking of the needles would account for the increased dispersing activity. The
increase of the anionic character caused by the adsorbed lignosulfonate causes repulsion between
particles, resulting in dispersion (Figure 2.15). The retardation effect of the sugars follows
different mechanisms that depend on the concentration and type of sugar and is discussed below
in the section on hydroxycarboxylic acids.

Cellulose Derivatives
Two cellulose polymers, hydroxyethyl cellulose (HEC) and carboxymethyl hydroxyethyl
cellulose (CMHEC), are used in well-cementing applications. Though HEC is not commonly
considered a retarder, it can extend thickening time by approximately two hours in a freshwater
slurry at a BHCT of 125F or less. Traditionally, the only cellulose that is considered a retarder
is CMHEC. This is largely due to the fact that CMHEC retards cement up to approximately
230F BHCT at the same concentrations as calcium lignosulfonate. CMHECs advantages
include:
It is less sensitive to cement quality than lignosulfonate. Formatted: Indent: Left: 1", Bulleted + Level:
1 + Aligned at: 0" + Tab after: 0.25" + Indent
It prevents gelation due to secondary reactions, for example, as a result of at: 0.25", Tab stops: 1.25", List tab + Not at
high free-alkali content and low gypsum content. 0.25"
It improves foam stability in foamed-cementing applications.
It contributes to slurry stability.
It inhibits condensation of vinyl polymers.
It improves fluid loss.
It does not affect long-term compressive strength.
It effectiveness is not shear-history dependent.

A perceived disadvantage of CMHEC as a retarder is its tendency to viscosify the slurry.

However, in applications where CMHEC is used as a retarder, the concentration requirement is
such that slurry viscosification is not a problem (Table 2.15). CMHEC can be used at
concentrations up to 0.7% without adding extra water to control slurry viscosity. Thereafter, to
achieve an equivalent slurry viscosity, the addition of 0.8 to 1.0 gal of water per sack of cement
is recommended for each percentage of CMHEC used.

Cellulose derivatives retard thickening time by forming an encapsulating film (adsorption) on

initially hydrated cement particles. The encapsulating film is composed of an already-hydrolyzed
cellulose polymer that further inhibits passage of water to the cement particle. A concentration in
excess of that required to encapsulate the particles increases the film density, which viscosifies
the aqueous phase and alters retardation time.

34
Hydroxycarboxylic acids. The hydroxycarboxylic acids are well known for their antioxidant
and sequestering properties that benefit cement slurry performance. The antioxidant property
improves the temperature stability of soluble compounds such as fluid-loss additives. The
sequestrant property of these compounds is the primary source of their retardation ability161.
Commonly used hydroxycarboxylic acids and their derivatives are citric acid, tartaric acid,
gluconic acid, glucoheptonate, and glucono-delta-lactone. The characteristic chemical formula
for hydroxycarboxylic acids is HOOC (CHOH)n COOH (Figure 2.20). The commonly used
hydroxycarboxylic acids are generally derived from naturally occurring sugars.

Numerous studies have been performed on the hydration of cement containing

hydroxycarboxylic acids.162-164 Hydrate phases absorb most of the product and only a small
amount is absorbed before the hydrated phases begin to form.165,166 The hydroxycarboxylic acids
accelerate the initial dissolution of both the aluminate and ferrite phases, causing rapid
precipitation of the AFt precursor gel on the cement particle surface (Figure 2.9). The low
permeability of the gel prevents further hydration, resulting in retardation. In addition, the
hydroxycarboxylic acids sequester the calcium during the growth of the C-S-H and CH.167 In the
case of the C-S-H, the morphology is altered by absorption of the hydroxycarboxylic acids into
the structure. This decreases the permeability of the C-S-H phase. A preference for absorption of
Al verses Ca has been observed. For effective retardation, it is preferred that the C-S-H and CH
precipitation be controlled, therefore, hydroxycarboxylic acids are more effective when used in
cement with a low aluminate and ferrite content. In such cements, less of the hydroxycarboxylic
acid is absorbed by the aluminate and ferrite phases, leaving more to react with the C-S-H and
CH phases. Because solution pH is critical to CH precipitation, the cement should contain a low
free-alkali content.

Organophosphonates. With a few exceptions, organophosphonates are the most powerful

retarders used in cement. However, these materials are not widely used in well cementing
applications because of
the low concentration required Formatted: Indent: Left: 0.04", Bulleted +
Level: 1 + Aligned at: 0" + Tab after: 0.25" +
difficulty of accurate measurement Indent at: 0.25"
sensitivity to concentration
The advantage of organophosphate retarders is in ultra-high temperature wells or in applications
where thickening times of 24 hours or greater are desired.

The mechanism of retardation is not well known. Organophosphonates are able to form
complexes with inorganic compounds, indicating that the complexation theory may apply.168 The
effectiveness of an organophosphonate retarder is dependent upon the number of phosphonate
groups and their position in the molecule (Figure 2.21).

Synthetic Retarders. The term synthetic retarder is a misnomer in that the previously discussed
retarding compounds are all, in effect, manmade. However, the term synthetic retarder has been
applied to a family of low-molecular-weight vinyl polymers. These retarders are based on the
same function groups as conventional retarders (i.e., sulfonate, carboxylic, or an aromatic ring).
The advantage of synthetic retarders is the ability to control the number and the relative position
of the pendent chain containing the function groups along the polymer backbone.

Several co-polymers are known to function as cement retarders. These include styrene sulfonate
co-maleic acid (SS-MA), 2-acrylamido-3-methyl propane sulfonate co-acrylic acid (AMPS-AA),
2-acrylamido-3-methyl propane sulfonate co-itaconic acid (AMPS-IA), or their alkali salts. In
35
general, SS-MA is also an active dispersant, depending on the amount of SS in the copolymer.
The AMPS-IA and AMPS-AA have a minor dispersing effect as compared to SS-MA, and are
weaker retarders. The AMPS-IA is a stronger retarder and dispersant than the AMPS-AA. All
three materials will contribute to fluid-loss control, but none are effective as a dispersant or fluid-
loss additive when used alone.

Inorganic Compounds. The retardation mechanism of inorganic compounds on cement

hydration is different than that for the previously discussed retarders. The inorganic compounds
commonly used as cement retarders are borax (Na2B4O7.10H2O) and other borates such as boric
acid (H3BO3) and its sodium salt, and zinc oxide (ZnO).

Borates are commonly used as an aid for high-temperature retarders at BHCTs of 300F (149C)
and above. At high temperatures, the borate is a less-powerful retarder than at lower
temperatures, however, it acts synergistically with other retarders such as lignosulfonates to
provide better retardation than either material can alone. The ratio between the retarder aid and
primary retarder varies depending on BHCT and the specific primary retarder. Borate has a dual
retardation effect in that it lowers the pH of the aqueous phase and is incorporated in the C-S-H
gel, reducing the gel permeability. It is an effective retarder when used alone. However, it is
detrimental to the performance of fluid-loss additives.

ZnO is a strong retarder when used alone. It is normally used for retarding chemically extended
cements (Section 2.11.3). Zinc compounds, such as ZnO, retard the alite phase more readily than
the aluminate phase. It is assumed that ZnO is converted to either Zn(OH)2 or
CaZn2(OH)6.2H2O, and is entangled within the C-S-H gel while simultaneously a fractional
amount is incorporated into the C-S-H gel.169,170 The combined effect is a reduction of the C-S-H
gel permeability. The precipitation of CaZn2(OH)6.2H2O also reduces the CH content in the
aqueous phase. The main advantage of ZnO as a retarder is that it will not inhibit the aluminate
hydration, resulting in improved effectiveness of chemical extenders. A disadvantage is the
potential for cross-linking with some fluid-loss additives.

Saturated saltwater. Water containing salt concentrations greater than 20% BWOW have a
retarding effect on cement (Figure 2.17). The retarding effect is caused by the formation of
gelatinous calcium monochloroaluminate (C3A.CaCl2.10H2O).171 The gelation is evident in the
thickening-time viscosity profile of saturated salt slurries by a sudden increase in Bearden units
of consistency that then levels off prior to set (Figure 2.22). Saturated salt slurries are useful for
cementing through salt domes. They also help protect shale sections from sloughing and heaving
during cementing and aid in preventing annular bridging and the lost circulation that could result.

2.11.3 Lightweight Additives/Extenders

Neat cement slurries prepared from API Class A, C, G or H cements using the amount of water
recommended in API Specification 10A13 will have slurry weights in excess of 15 lbm/gal. In
many parts of the world, severe lost-circulation and weak formations with low fracture gradients
are common. These situations require the use of low-density cement systems that reduce the
hydrostatic pressure of the fluid column during cement placement. Lightweight additives (also
known as extenders) help reduce the weight of the slurry. They include:
physical extenders clays and organics Formatted: Indent: Left: 1", Bulleted + Level:
1 + Aligned at: 0" + Tab after: 0.25" + Indent
pozzolanic extenders at: 0.25", Tab stops: Not at 0.25"
chemical extenders
36
 gases

Any material with a specific gravity lower than that of the cement will act as an extender. These
materials, in general, decrease the density of cement slurries by one of three means. (1) The
pozzolanic and inert organic materials have a lower density than cement and can be used to
partially replace cement, therefore lowering the density of the solid material in the slurry. (2)
Physical and chemical extenders not only have a lower density but also absorb water, thus
allowing more water to be added to the slurry without producing free fluid or particle
segregation. (3) Gases are used to produce foamed cements that have exceptionally low density
with acceptable compressive strengths. In many lightweight slurries, a combination of materials
are used. For example, pozzolanic and chemical extenders can be used with physical extenders
and/or gases. Pozzolan slurry designs almost always incorporate bentonite, and gases are
generally used with a chemical extender to stabilize the foam. Lightweight additives also
increase the slurry yield and can result in a more economical slurry.

2.11.3.1 Physical Extenders

Physical extenders are particulate materials that function as cement extenders by increasing the
water requirements or by reducing the average specific gravity of the dry mix. Two general
classes of materials fall into this category: clays and inert organic materials. The most commonly
used clay material is bentonite, although attapulgite is also used. The commonly used inert
organic materials are perlite, gilsonite, ground coal, and ground rubber.

Bentonite (gel) is a colloidal clay mineral composed predominantly of sodium montmorillonite

(NaAl2(AlSi3O10).2OH). Physically, it is composed of three silica platelets bound together by a
metal ion. It has the unique property of imparting viscosity to cement slurries by dissociation of
the silica platelets absorbing large quantities of water. It has been reported that in freshwater, the
dissociated volume of the silica platelets can result in a 10-fold increase in the original particle
volume. In salt-containing waters, the effectiveness is reduced. The montmorillonite content of
bentonite is the controlling factor in its effectiveness as an extender, hence, it is one of two
extenders that are covered by an API specification. Bentonite can be added to any API class of
cement and is commonly used with other extenders. Bentonite is used to prevent solids
separation, reduce free water, reduce fluid loss, and increase slurry yield.

Bentonite is typically used at concentrations of 1 to 16 % BWOC (Table 2.16). It may be dry-

blended with the cement or prehydrated in the mixing water. The effect of 1% BWOC
prehydrated bentonite is equal to about 3.5 % BWOC dry-mixed, but the yield point is much
higher. The mixing method used to prehydrate bentonite will affect the time required for
complete dissociation of the silica platelets. Prehydration takes approximately 30 minutes in a
low-shear mixer, or approximately 5 minutes in a high-shear mixer. For best results, the
prehydrated bentonite-water mixture should be used for mixing the cement slurry shortly after
prehydration is completed. Allowing the bentonite to prehydrate for an extended time can reduce
its effectiveness, increasing solids separation and free water. Because of changes in slurry
performance properties (i.e., thickening time and compressive strength), prehydration is not a
direct substitute for dry-blending. Laboratory testing is required to determine the effect of
substitution of prehydration for dry-blending.

Cement slurries containing high bentonite concentrations (i.e., 8 to 25% BWOC) frequently
require additional additives for controlling slurry viscosity and thickening time. Calcium
lignosulfonate is commonly used as a dispersant and retarder in high-gel cements to reduce
37
slurry viscosity. Concentration requirements vary, depending on the bentonite content in the
slurry and whether the bentonite was dry-blended or prehydrated. The concentration of
lignosulfonate is lower for prehydrated bentonite than for dry blends. When bentonite reacts with
lignosulfonate, a portion of the lignosulfonate is consumed on the platelet surface and is no
longer available for slurry retardation. Therefore, the lignosulfonate concentration required to
achieve retardation is higher for bentonite slurries than for non-bentonite slurries. Slurries
containing 8 to 25% BWOC and lignosulfonate are referred to as modified cements (Table 2.17).
Sodium chloride is also used in high-bentonite cements. Sodium chloride inhibits the
dissociation of bentonite and is an effective dispersant and accelerator. Sodium chloride and
lignosulfonate can be used together to formulate 12 to 16 % BWOC cement slurries with reduced
water requirements (Table 2.18).

Attapulgite (salt gel) is a more effective extender than bentonite in seawater or high-salt slurries.
Attapulgite, (Mg,Al)2[OH/Si4O10].12H2O, is composed of clusters of fibrous needles that require
high shear for dispersion in water. It produces many of the same effects as bentonite, but it does
not reduce fluid loss because of the higher permeability generated in the filter cake by fibrous
needles as compared to platelets. Because the fibers are similar to asbestos fibers, its use has
been prohibited in some countries. Granular forms may be permitted as a replacement.

Expanded perlite is a siliceous volcanic glass that is heat-processed to form a porous particle
that contains entrained air. It is a highly buoyant product that requires the addition of 2 to 6 %
BWOC bentonite to prevent separation from the slurry. Because of its low crush strength, the
water requirement for perlite-containing slurries must be increased to allow for slurry
compressibility under downhole conditions. Volume loss must also be taken into effect in fill-
volume calculations.

Gilsonite is an asphaltic material, or solid hydrocarbon, found only in Utah and Colorado. It is
one of the purest naturally occurring bitumens. Because of gilsonites low density (1.07 g/cm3),
it can be used to prepare slurry densities as low as 11 lb/gal at a concentration of 100 lb/sk of
cement. An organic material, gilsonite is highly buoyant and will float out of the slurry unless
inhibited. A concentration of 2 to 6% bentonite is commonly added to prevent separation from
the slurry. Other disadvantages of gilsonite include a relatively low softening point, ~240F
(116C), and melting point, ~383F (195C), and cost. The main advantage is its low water
requirement that results in improved compressive strengths compared to other physical
extenders. It is also used for lost circulation.

Crushed coal is used for the same purposes as gilsonite (i.e., for light weight and lost
circulation). It is commonly used at concentrations up to 50 lb/sk of cement. Its density is
slightly higher (1.3 g/cm3), requiring a slightly higher water content. The addition of bentonite to
prevent separation is normally not required but bentonite is advantageous for achieving lower
slurry densities. Because of its very high melting point, ~1000F (538C), crushed coal is useful
for high-temperature applications.

Ground rubber is a low-cost alternative to gilsonite and may be used in similar applications.
The density of ground rubber is slightly higher (1.14 g/cm3). The physical properties are more
variable than gilsonite and are dependent upon the material source. A major advantage of ground
rubber is its low cost.

38
 2.11.3.2 Pozzolanic Extenders
Several pozzolanic materials are available for producing lightweight cement slurries. These can
be either natural or artificial and include fly ash, diatomaceous earth, microsilica, metakaolin,
and granulated blast furnace slag. (See Section 2.9.1). In comparison with other additives,
pozzolanic materials are usually added in large volumes. Fly ash, for example, can be mixed with
cement in ratios of fly ash to cement that range from 20:80 to 80:20 based on an equivalent
sack weightthat is, where a sack of fly ash has the same absolute volume as that of a sack of
cement. Pozzolanic materials have a lower specific gravity than cement, and this lower specific
gravity gives a pozzolanic Portland cement slurry a lower density than a Portland cement slurry
of similar consistency (Table 2.8). Pozzolanic materials react with the CH formed on hydration
of alite and belite in cement to form additional C-S-H binder, thereby increasing compressive
strength and reducing permeability. Depending on their density, pozzolanic cements also tend to
give a set cement that is more resistant to attack by formation waters.

Fly ash is by far the most widely used of the pozzolanic materials. According to ASTM C618,
there are two types of fly ash, Class F and Class C: Class N refers to natural pozzolanic
materials. There is however, a need for a third category based on the performance of different fly
ashes. ASTM C618 classifies fly ash on the basis of combined percentages of SiO2 + Al2O3 +
Fe2O3; Class F having a minimum of 70 % and Class C, 50%.172 In reality, there is a much
greater relationship between CaO content and performance.173 The CaO content ranges from 2 to
3% to 30% by weight of the fly ash. The true Class F fly ash has less than 10% CaO, whereas
a true Class C has more than 20% CaO. Fly ashes with 10 to 20% CaO behave somewhat
differently than the true Class F or Class C. Fly ashes are generally composed of amorphous
glassy particles that are spherical in nature. Mineral phases are often also found embedded close
to the surface, and these make up the remaining composition of the fly ash. The primary
differences between the Class F and Class C fly ashes are:

Class F fly ash (CaO <10%) is composed of 70 to 90% of vitreous silica-rich aluminosilicate
glassy spheres174, the remainder being composed essentially of the embedded mineral
phases.175 Only a few different mineral phases are present (Table 2.19), and they tend be
slow to react in cement up to 302F (150C). This fly ash behaves only as a pozzolan in that
the siliceous glassy phase or crystalline quartz reacts with the CH from cement hydration to
form C-S-H. Specific gravities typically range from 1.6 to 2.8.
Class C fly ash (CaO >20%) is composed of calcium aluminosilicate glassy spheres174 that
contain mineral phases not only embedded in them but also coating the surface.175 Many
mineral phases can be present (Table 2.19), and some, such as the C3A, are cementitious in
behavior. Fly ashes with high CaO content tend to be very reactive both in terms of the glass
phase and as the cementitious mineral phases. This fly ash is both pozzolanic and
cementitious in behavior. Some are so cementitious that they can harden almost immediately
upon mixing with water. Specific gravities commonly range from 2.4 to 2.8.
Fly ash with CaO content of 10 to 20% is composed of glassy spheres that range from
vitreous silica to calcium aluminosilicate in composition.174 The mineral phase composition
varies and, in some cases, is similar to that of a Class F while in others to a Class C fly ash.
In general, the higher the CaO content of the fly ash, the greater the initial early strength,
though mineral composition, glass composition, and fineness can also contribute to a greater
or lesser extent. A high proportion of cementitious phases can lead to increased viscosity and
ultimately cause gelation. The fly ashes in this category can behave as a pozzolan only or
have both pozzolanic and cementitious properties.

39
The ASTM standard specification Class F fly ash is the most common used in oilwell cementing,
and is covered by the API specifications.13 The major advantages of the Class F fly ash is its low
cost and abundance worldwide. The performance characteristics of a Class F fly ash vary little
from batch to batch from a given source. However, the performance of fly ash from different
sources can vary considerably because the composition can vary from the true low CaO to 10 to
20% CaO. As a result, different sources of Class F fly ashes should be tested before use. Specific
gravities also must be determined. Some power plants produce Class F fly ashes with a high
carbon content because of poor burning, and these should be avoided as they can cause severe
gelation problems. A disadvantage of Class F fly ash cements is that early strength development
is usually significantly less than that of a neat cement mixed at the same density because the fly
ash initially behaves as an inert filler. The increased strength resulting from the pozzolanic
reaction does not normally occur until after 12 hours of hydration. The early compressive
strength can be improved by the addition of accelerators, Section 2.11.1. Bentonite is normally
added to Class F fly ash slurries at concentrations of 2 to 12% depending on slurry density to
prevent free fluid development. The resulting increase in porosity, however, is also associated
with low compressive strengths.

The use of Class C fly ash as an extender for well cementing is relatively limited. This is due, in
part, to the limited availability of Class C fly ash and to the considerable performance variability
that exists not only between sources but also to a large extent between batches from a given
source. Class C can give very good initial compressive strengths and, if it is highly cementitious,
it can be used as the principal component of the cement. The variability with the potential for
rapid set and gelation, however, makes the use of such systems problematic. Performance should
always be verified before use.

Microspheres are used when slurry densities from 8.5 to 11 lb/gal are required. Microspheres
are hollow spheres obtained as a byproduct from power-generating plants or are specifically
formulated. The byproduct microspheres are essentially hollow fly-ash glass spheres. They are
typically present in Class F fly ashes, but usually in small amounts. However, they are obtained
in substantial quantities when excess fly ash is disposed of in waste lagoons. The low-density
hollow microspheres float to the top and are separated by a flotation process. These hollow
microspheres are composed of silica-rich aluminosilicate glasses typical of fly ash and are
generally filled with a mixture of combustion gases such as CO2, NOx, and SOx. The synthetic
hollow microspheres are manufactured from a soda-lime borosilicate glass, are formulated to
provide a high strength-to-weight ratio, and are typically filled with nitrogen. The synthesized
microspheres provide a more consistent composition and exhibit better resistance to mechanical
shear and hydraulic pressure. Microspheres range in size from 20 to 150 m, with the wall
thickness varying from 0.5 to 2.0 m. The specific gravity of the byproduct microspheres is
approximately 0.6, whereas the specific gravity of the synthetic microspheres ranges from 0.12
to 0.6. Typically, synthetic microspheres have a specific gravity in the range of 0.4 to 0.6 and are
used in oilwell cementing. The spheres with very low specific gravity tend to break and collapse
when exposed to pressure or shear. In fact, this phenomenon, known as crushing, occurs for all
microspheres when the resistance of the individual microsphere is exceeded. Resistance to
crushing depends on the size, thickness, and composition of the microsphere (Figure 2.23).
Byproduct microspheres, in general, have crush strengths of 4,000 to 5,000 psi, whereas
synthetic microspheres in the specific gravity range of 0.4 to 0.6 can survive from 4,000 to
10,000 psi, depending on the category of microsphere used. However, because of the relative
expense of synthetic microspheres, byproduct microspheres are more commonly used unless
circumstances dictate otherwise.

40
The primary disadvantage of most sources of microspheres is their susceptibility to crushing
during mixing and pumping and when exposed to hydrostatic pressures above the average crush
strength (Table 2.20). This can lead to:
b) increased slurry density Formatted: Indent: Left: 1", Bulleted + Level:
1 + Aligned at: 0" + Tab after: 0.25" + Indent
c) increased slurry viscosity at: 0.25", Tab stops: Not at 0.25"
d) decreased slurry volume
e) premature slurry dehydration

Crushing effects can be minimized by the choice of microspheres. These effects can be predicted
and can be taken into account in slurry design calculations to produce a slurry with the
characteristics required for the well conditions. Lightweight systems incorporating microspheres
can provide excellent strength development and can help control fluid loss, settling, and free
water. Systems containing a high percentage of microspheres can exhibit poor mechanical
properties because of the brittle nature of the microspheres. It is important to ensure that the
microspheres do not separate from the slurry during blending or transportation to the well site
and that the cement is thoroughly blended before use. Separation of the microspheres from the
slurry during mixing will result in varying densities and thus, erratic performance of the set
cement.

Both the byproduct and synthetic microspheres are pozzolanic and will ultimately react with the
CH produced from the cement to produce additional C-S-H. The volume of microspheres,
however, is usually greater than that of the CH produced from the cement and, in most cases, the
spheres will have only a surface coating of reaction product. The compressive strength is thus
dependent upon the final microsphere concentration.

Microsilica, also known as silica fume, is a finely divided, high-surface-area silica that can be
obtained as a liquid or powder. In powder form, it can be either in its original state, densified, or
pelletized. The average particle size is 0.1 to 0.2 m, approximately 100 times smaller than that
of an average cement particle. In relation to cement, which has a specific surface area of 1500
m2/kg as determined by N2 adsorption, the specific surface area for microsilica ranges from
15000 to 25000 m2/kg. The true density is above 2000 kg/m3, though the bulk density of the
original freshly filtered powder is about 240 to 300 kg/m3. This makes the material difficult to
handle unless it is densified or pelletized. The bulk density of the densified microsilica is
approximately 400 to 500 kg/m3. Microsilica typically has a specific gravity of approximately
2.2.

Microsilica is composed primarily of vitreous silica, and has a SiO2 content of 85 to 95%, which
makes it considerably purer than the other pozzolanic materials. KCl, quartz, metallic iron, and
iron silicide are the most common crystalline impurities detected. Because the raw materials used
to make silicon or ferrosilicon are quite pure, the microsilica collected from a given furnace is
usually of consistent composition. In some cases, however, the microsilica collected at one
source may be from different furnaces producing different alloys, so it can be a mixture
composed of variable chemical compositions and pozzolanic activity.

In comparison with the other pozzolanic materials, microsilica is very reactive because it has a
high SiO2 content, is amorphous and, more importantly, is extremely fine. Microsilica reacts
almost immediately and appears to act as a nucleation site for precipitation of the CH as it is
released during the cement hydration, causing formation of numerous small crystals of CH176
41
that then react to form a cementitious material. The pH of microsilica slurries is also significantly
lower177 and some experimental evidence suggests that the cementitious material is not a C-S-H
as formed with other pozzolanic materials but a microcrystalline hydrosilicate containing alkalis
similar to alkali-silica gels.178 The alite reaction appears to be accelerated179 and the Ca/Si ratio
of the C-S-H formed from hydration of the cement decreases markedly as the content of
microsilica increases.

Microsilica particles are also considered to impart beneficial physical properties to the slurry.
Because of their fineness, the particles are believed to fill the voids between the larger cement
particles, resulting in a dense, solid matrix, even before any chemical reaction between the
cement particles occurs. Microsilica also tends to improve rheological properties because the tiny
spheres can act as very small ball bearings and can displace some of the water present between
the flocculated cement grain, thereby increasing the amount of available fluid. Concentrations of
microsilica can range from 3 to 30% BWOC, depending on the slurry and properties required.

The physical and chemical properties of the microsilica make it very useful for a variety of
applications other than as an extender. These include
compressive-strength enhancement for low-temperature lightweight cement Formatted: Indent: Left: 1", Bulleted + Level:
1 + Aligned at: 0" + Tab after: 0.25" + Indent
thixotropic properties for squeeze cementing, lost circulation, and gas at: 0.25", Tab stops: 1.25", List tab + Not at
migration 0.25"
providing some fluid-loss control

The one disadvantage of microsilica is the cost. Originally, it was considered to be a waste
product, but with its increased usage in the construction industry over the last decade, it has
become more of a specialty chemical. Also, with fluctuations of supply and demand, a constant
supply from a good source is not always assured.

Diatomaceous earth is a natural pozzolan composed of the skeletons of microorganisms

(diatoms) that were deposited in either fresh water or seawater. The skeletons consist of a
hydrated amorphous silica, opal (SiO2.xH2O), that contains up to 15% water. Diatomaceous
earths used as cement extenders are ground to a particle size of 3.0 to 3.5 m2/kg, similar to
cements, though finenesses above 10 to 12 m2/kg are achievable. The water content of
diatomaceous earths can also be reduced to <3% by drying. The specific gravity is generally 2.0
to 2.2.

Crude diatomaceous earths are often contaminated with clay minerals or other minerals such as
quartz and feldspars, and the opal content can range from 25 to 100%. It is the presence of the
opal that gives the diatomaceous earth its pozzolanic behavior. When used as pozzolans,
diatomaceous earths containing clays can be made more effective if they are calcined at
temperatures between 1400 and 1832F (760 to 1000C). This converts the crystalline structure
of the clays to an amorphous or disordered structure, making them highly pozzolanic.

The complex internal structure of the diatoms creates an extremely high specific surface area,
usually in the range of 1700 to 45000 m2/kg. This causes the water requirements to be very high.
However, diatomaceous earth can be used in large amounts up to 40% by weight of cement
without an adverse increase in viscosity, depending on the amount of water used. Typically,
extended slurry densities as low as 11.0 lb/gal are achievable (Figure 2.24).

42
Although diatomaceous earth costs more, there are several advantages to using it instead of clays
or fly ash. As a processed material, the composition, particle size, and specific surface area of
diatomaceous earth can be selected to achieve the desired slurry properties. It is effective in
formulating stable lightweight slurries at high temperatures. Since diatomaceous earth typically
contains a minimum of 86% silica, it is an effective substitute for a portion of the silica sand or
silica flour for the prevention of strength retrogression (Sections 2.10.3 and 2.11.8). The small,
porous particles also impart insulation properties to set cements that are similar to those of
foamed cement. Both of these properties are particularly beneficial for cementing geothermal or
steam flood wells (Figure 2.25). Diatomaceous earths are less susceptible to deterioration by
sulfate ground waters than fly ash blends, and even more so than clays, because of their smaller
particle size. Furthermore, unlike other extenders, compressive strength is improved by
increasing the concentration used. Diatomaceous earth cements also have excellent perforation
characteristics, including greater penetration depth and much less shattering. In general,
diatomaceous earth is a more effective extender than other materials when used below 13 lb/gal
or at temperatures greater than 140F (60C) (Table 2.21). If used at temperatures less than
140F (60C), calcium chloride is recommended for acceleration.

2.11.3.3 Chemical Extenders

Several materials are effective as chemical extenders. In general, any material that can
predictability accelerate and increase the concentration of the initial hydration products is
effective as a chemical extender. Materials of this type must contain anions or cations that are
common to the cement itself or that readily substitute for these ions in the hydration reactions.
Therefore, non-acidic compounds containing silicate, sulfate, or aluminum are effective as
chemical extenders.

Sodium silicate is the most commonly used chemical extender for cement slurries. Sodium
silicate is five to six times as effective as bentonite in equivalent concentrations. Unlike the
physical or pozzolanic extenders, sodium silicate is highly reactive with the cement. Sodium
silicate reacts with the Ca2+ ions in the aqueous phase to form additional C-S-H gel (Figure
2.11). The increase in C-S-H is sufficient to adsorb large quantities of water, thereby increasing
the slurry viscosity and accelerating further C-S-H formation, resulting in a reduction in
thickening time. The efficiency of sodium silicate is improved by catalyzing the reaction with
CaCl2 or by mixing with seawater.

Sodium silicate is available in both dry and liquid forms, making it readily adaptable to onshore
and offshore applications. The solid form is sodium metasilicate (Na2SiO3), which is typically
dry-blended with cement at a concentration of 1 to 3.5% BWOC at densities of 14.2 to 11.5
lb/gal. It is less effective if dissolved directly in the mix water unless CaCl2 is dissolved in the
water first. If a liquid system is desired, the liquid form of sodium silicate is recommended. The
liquid form of sodium silicate is a silicate polymer containing three to five silicate molecules in
the chain (Na2O.(35)SiO2). Sodium silicates containing more than five silicate molecules are
not effective as chemical extenders because of their reduced solubility. Liquid sodium silicate is
normally used in seawater applications at a concentration of 0.1 to 0.8 gal/sk of cement at
densities of 14.2 to 11.5 lb/gal. The two main advantages of sodium silicates as extenders are
their high yield and low concentration of use. Their accelerating effect, however, limits their
application at lower temperatures, typically < 125F (52C) BHCT. They can be used at higher
temperatures with the addition of retarders, although their effectiveness as an extender is reduced
because of the inhibition of C-S-H formation. Sodium silicates are also known to reduce the
effectiveness of most fluid-loss additives. Typical performance data is presented in Figure 2.26.

43
Gypsum - The use of gypsum(CaSO4.2H2O) has been discussed previously (Section 2.9.3). The
hemihydrate form of calcium sulfate (CaSO4.0.5 H2O) often though incorrectly referred to as
gypsum is typically used as an extender. Concentrations of 15 % BWOC or less hemihydrate are
typically used in thixotropic slurries for treating severe lost-circulation problems or applications
in which expansion properties are desired to improve bonding. In either case, effectiveness is
improved if the C3A content is 5 wt.% or greater because expansion properties are inhibited
below this value. Class A cement with C3A content >10 wt.% can be used to prepare 50:50 fly
ash slurries. Slurry compositions typically used for lost-circulation applications [BHCT 125F
(52C)] contain 8 to 12 % BWOC gypsum with good expansion properties (0.2 to 0.4%). For
improved bonding applications where increased expansion (0.4 to 1%) is desired, NaCl is used
(% BWOW 10).

Foamed Cement. Slurries ranging in density from 4 to 15 lb/gal can be prepared with foamed
cement. Foamed cement is created when a gas, usually nitrogen, is injected at high pressure into
a base slurry that incorporates a foaming agent and foam stabilizer. Nitrogen gas can be
considered inert, and does not react with or modify the cement hydration product. When properly
mixed and sheared, foamed cements contain very small discrete bubbles that will not coalesce or
migrate. Such cements produce a low-density matrix with low permeability and relatively high
strength. The base slurry is specially formulated so that it is homogenous, and cement admixes
must be chosen carefully. The density of the base slurry can range from lightweight to heavy,
depending on the application. Foamed cement can be used at temperatures from 28 to 600F (82
to 1112C). The quality of the foamed cement slurry is dependent on well depth, temperature,
and desired density downhole (Table 2.22).

Advantages of the foamed structure include

enhanced mud removal
good ductility
expansive properties
prevention of gas migration
fluid-loss control
stabilization of set cement at high temperatures
insulation properties
compatibility with non-Portland cements

The disadvantage of foamed cement is the need for specialized cementing equipment for both
field application and laboratory testing.

Under special circumstances, compressed air can be used for low-pressure or shallow
applications. Obtaining compressed air at a sufficient volume, rate, and pressure to provide the
level of shear required for a stable foam can, however, be a challenge.

2.11.4 Weighting Agents

Weighting agents or heavyweight additives are used to increase slurry density for control of
highly pressured wells. Weighting agents are normally required to achieve densities greater than
17 lb/gal when dispersants or silica are no longer effective. The main requirements for weighting
agents are that

44
 the specific gravity is greater than the cement
the particle size distribution is consistent
they have a low water requirement
they are chemically inert in the cement slurry
they do not interfere with logging tools

Through experience, a preferred specific gravity range of 4.5 to 5.0 g/cm3 has been established.
This provides a balance between the particle size required to inhibit settling and the increase in
water requirement necessary for wetting the particle surface while maintaining the viscosity
within an acceptable range. Weighting agents must be chemically inert so that they do not affect
other properties of the cement, such as thickening time.

Hematite (Fe2O3) is the most commonly used weighting agent. Hematite is a brick-red, naturally
occurring mineral with a dull metallic luster, depending on the amount of soluble iron oxide. It
contains about 70% iron. The specific gravity of hematite ranges from 4.9 to 5.3, depending on
purity, and it has an approximate Mohs hardness of 6. The average specific gravity of hematite
typically reported is 5.0 g/cm3, but its specific gravity will vary, depending on the source.
Though it is available in various particle-size grades, the fine-grind material typically used is
90% < 45m. Hematite is an efficient weighting agent for most applications. In higher-density
applications where the viscosity produced by the fine grade is unacceptable, mixtures of the three
different grades may be used. The fine grade is typically used to prepare slurries with densities
up to 19 lb/gal. Coarser grades (70% < 45m) can be used to prepare slurries of comparable
viscosity up to 22 lb/gal. A disadvantage to hematite is the presence of variable amounts of
soluble iron oxide (rust FeO2) that will affect the aluminate and ferrite reaction as well as
precipitate Fe(OH)2 3 gel from solution when exposed to caustic conditions, resulting in higher
slurry viscosities during mixing. It also affects the viscosity contribution of several of the
commonly used fluid-loss agents.

Ilmenite (FeO.TiO2) is not as commonly used as hematite, although it has some distinct
advantages. Ilmenite is a black to dark brownish-black, naturally occurring mineral with a
submetallic luster that contains about 37% iron. It resembles magnetite in appearance, but has
only a slight magnetic character. The specific gravity ranges from 4.5 to 5, depending on the
purity, and it has a Mohs hardness of 5 to 6. The average specific gravity typically reported is
4.7, but will vary, depending on the source. The commonly used particle size is fairly coarse,
compared to other weighting agents. As a result of this, it imparts very little viscosity to the
slurry, allowing higher densities to be achieved at lower concentrations than hematite, making
slurry densities of 22 lb/gal much easier to achieve. Because of its lower iron content and the
excellent resistance to atmospheric corrosion from the titanium component of the ilmenite,
soluble oxides are not a problem; therefore, slurry viscosity is generally not an issue. A
disadvantage of ilmenite is that it is more sensitive to the particle-size distribution within the
slurry, and thus is more susceptible to sedimentation. This is easily addressed by adding a small
percentage of microsilica or by using silica flour rather than silica sand in high-temperature
slurries.

Hausmannite (Mn3O4) is increasingly used because of its unique properties that address many
of the disadvantages of other weighting agents. Hausmannite is a dark brownish-black material
that is a byproduct mineral from the processing industry. The specific gravity range or Mohs
hardness has not been well established. The average specific gravity typically reported is 4.3
g/cm3, and the average particle size is 5 m. There are several unique advantages to the use of
45
this material. Because of its particle size and unique wetting characteristics, the material can be
suspended in the mix water at up to 40 wt.% with a minimum of agitation, providing a liquid
weighting agent. Because the average particle size of hausmannite is much smaller than that of
cement, it can fit within the cement pore matrix, displacing entrained water, which results in a
lower viscosity and significantly more stable slurry. The main disadvantage is that it is not
readily available in all geographic regions, and the additional shipping cost can make it cost-
prohibitive.

Barite (BaSO4) is not normally used in cementing as a weighting agent because of its high
surface area and high water demand. It is a soft, light gray, naturally occurring nonmetallic
material. The specific gravity ranges from 4.0 to 4.5, depending on purity, and it has a Mohs
hardness of 2.5 to 3.5. The average specific gravity typically reported is 4.23 g/cm3, but its
specific gravity will vary according to the source. The particle size distribution is variable;
however, according to API specifications, 30 wt.% can be less than 6m and, in practice, 90
wt.% is less than 45m. The high water demand of barite results in higher concentration
requirements than any other weighting agent. This, plus the additional water required for
viscosity control, further reduces the cement/water ratio, further decreasing compressive
strength. Barite can be used to achieve slurry densities up to 19.5 lb/gal.

2.11.5 Dispersants

Dispersants, also known as friction reducers, are used extensively in cement slurries to improve
the rheological properties that relate to slurry flow behaviour. Dispersants are used primarily to
lower the frictional pressures of cement slurries while they are pumped into the well. This
practice effectively
reduces the pumping rate necessary to obtain turbulent flow for specific well Formatted: Indent: Left: 0.5", Bulleted +
Level: 1 + Aligned at: 0" + Tab after: 0.25" +
conditions Indent at: 0.25", Tab stops: 0.75", List tab +
reduces the surface pumping pressures and horsepower required to pump the cement Not at 0.25" + 1"
into the well
reduces pressures exerted on weak formations, possibly preventing circulation losses

Dispersants can also be used to give slurries having high solids-to-water ratios and good
rheological properties, allowing engineers to design high-density slurries (up to 17 lb/gal)
without a weighting additive (Table 2.23). Dispersants can also be used to design low-density
slurries in which the high-solids contents include lightweight extenders.

The charges produced on the surface of cement particles as they hydrate, coupled with the ions
present in the solution, cause electrostatic interaction between the particles and lead to
flocculation. This results in the formation of a weak structure throughout the slurry, and causes it
to behave as solid. A certain amount of shear stress is required to break the structurethe true
yield valueand allow the cement to flow. Cement-flow properties are complicated and do not
follow ideal Newtonian fluid behavior. Several models (Power Law, Bingham Plastic, Herschel-
Bulkley, Casson) define the effect of flow for a cement slurry. These relate the shear stress (force
per unit area causing the fluid to flow) to the shear rate (change in velocity occurring between
the applied force and a fixed element of fluid) (Figure 2. 27). Plotting the shear stress to shear
rate gives the plastic viscosity (slope of the line) and yield value (intercept) (Figure 2.28).
Because of the limitations of experiments, however, the yield value based on the models is not
equivalent to the true yield value. Furthermore, the slope of the line is typically not linear and a
46
measurement of apparent viscosity, obtained from the shear stress-to-shear rate ratio, is more
appropriate. The apparent viscosity is not a constant value, and for a cement slurry, it usually
decreases with shear stress. As the cement breaks up under the applied shear stress, the
flocculated structure breaks up into agglomerates of particles. It is the movement of these
particles in the base fluid that gives rise to the plastic or more correctly applied viscosity of the
slurry. The rheological properties of a cement slurry are dependent on many factors:
specific surface area of the cement Formatted: Indent: Left: 1.5", Bulleted +
Level: 1 + Aligned at: 0" + Tab after: 0.25" +
water-to-cement ratio Indent at: 0.25", Tab stops: Not at 0.25"
surface composition of the hydrating cement grain
mixing and conditioning procedure
time after mixing
temperature (hydration or test)
testing procedures

In the past, the effects and development of dispersed slurries were most often studied using the
simpler Power Law or Bingham Plastic models, with data usually represented in terms of plastic
viscosity and yield value. Essentially, dispersants decrease both the plastic viscosity and the
yield value of a cement slurry.180-182 In general, dispersants are effective only over given
concentration ranges that are dependent on the type of dispersant, the cement, and the slurry
design used. Usually, a minimum concentration of dispersant (Cmin) is required before any
notable dispersion effect occurs and an optimum concentration (Copt) establishes the point above
which viscosity and yield values reach a minimum and level off (Figure 2.29).182 Addition of
dispersant above the optimum concentration can be detrimental in causing particle segregation
and free-fluid problems.

Dispersants, also known as superplasticizers in the concrete industry, have been extensively
studied.183-186 It is generally agreed that dispersants minimize or prevent flocculation of cement
particles because the dispersant adsorbs onto the hydration cement particle, causing the particle
surfaces to be negatively charged and repel each other (Figure 2.15). Water that would have
otherwise been entrained in the flocculated system also becomes available to further lubricate the
slurry. Steric hindrance that would prevent particle-to-particle interactions is also possible when
high-molecular-weight dispersants are used.

Additives that disperse cement slurries appear to contain more than one polar group in the
molecule, and many are polymeric. Sulfonated polymers are most commonly used as cement
dispersants and include naphthalene sulfonate, melamine sulfonate, styrene sulfonate, phenol
sulfonate,187 and ketone sulfonate.188 Lignosulfonates, low-molecular-weight polymers such as
hydoxylated polysaccharides, and non-polymeric chemicals, in particular hydroxycarboxylic
acids, are good dispersing agents; however, because of their strong retardation effects, they are
not usually used as dispersants. It is not clear why the sulfonated polymers have an enhanced
dispersing power, but it probably relates to their weaker retarding ability, allowing them to be
used in higher concentration. The weaker retardation may result from their preferential reactivity
toward aluminate and ferrite phases.

The most frequently used of the sulfonated polymers are the salts of sulfonated naphthalene
formaldehyde condensate and, to a lesser extent, the salts of sulfonated melamine formaldehyde
condensate and polystyrene sulfonates.

47
Sulfonated naphthalene formaldehyde condensate (SNFC, PNS, NSF or SNF) is available as
a calcium and/or sodium salt, and can be obtained in both solid and liquid form. The commercial
liquid form typically has a solids content of approximately 40%. The degree of
condensation/polymerization is maintained between batches by controlling the ratio of
formaldehyde to naphthalene sulfonic acid, reaction temperature, and reaction time. At present,
no technique adequately characterizes the degree of polymerization or branching, if any, and
average molecular weight or molecular weight distributions reported are at best, approximations.
The repeating unit for SNFC is shown in Figure 2.30, and estimates indicate that chains
comprise 10 or fewer units in length.189 Differences in SNFC samples appear to have minimal
effect on dispersive properties (Figure 2.31).

The SNFC adsorbs on the positive sites of the hydrating cement particle and gives the resulting
particle a negative charge (Fig 2.15). Dispersion is then caused by repulsion between particles.
Some of the SNFC is adsorbed by the initially formed C-S-H; however, the adsorbed layer seems
to have minimal effect on permeability, and hydration processes proceed with only a slight
retardation in setting. The effect of the SNFC on the aluminate and ferrite phase hydration is
much more significant, and a surface chemical or chemical interaction seems to occur between
the aluminate, ferrite, sulfate, and SNFC. It would appear that the sulfonate groups of the SNFC
replace the sulfate from gypsum in the reaction (Figure 2.9), and poisons the growth of the AFt
needles from the precursor gel.87,190,191 The fact that the SNFC does not accelerate formation of
the gel, yet prevents growth of the AFt needles, further enhances its dispersive properties.
Alkalis also seem to play a significant role in the adsorption properties of the SNFC.87 The
dispersive ability of SNFC is highly dependent on the cement used (Figure 2.29) and relates to a
certain extent on its fineness and its aluminate and ferrite contents. 173,176,183,184

The benefits of using SNFC is that improved rheological properties can be obtained and slurries
can be pumped with reduced frictional pressures. SNFC also allows higher solid-to-water ratio
slurries to be designed with improved properties. The primary disadvantage is that if too much
dispersant is added, particle segregation and free fluid development can occur. Initially, this is
caused by the particles having sufficient opposing charge to cause all the particles to repel each
other, causing the particles to separate. Further increase in dispersant concentration causes
flocculation. This results from an increase in the ionic strength of the aqueous phase (Na+ or Ca2+
salts from the dispersant plus the charge surface of the dispersant) and increases pore-solution
viscosity (resulting from the intrinsic viscosity of the SNFC polymer), causing cohesion of the
particles with the dispersant /salts acting as glue. Similar effects also occur with high-NaCl salt
slurries.

Sulfonated melamine formaldehyde condensate (SMFC, MSF or SMF) is available as a

powder or as a liquid with 21% solids content. The average molecular weight is dependent
primarily on the polymerization time. Average molecular weights appear to range from 1200 to
44000,194 with molecular weights of approximately 30000189 having the most effective dispersion
but, as with SNFC, the validity of actual values is questionable. The structure of the repeating
unit for SNFC is shown in Figure 2.30. The SMFC behaves similarly to the SNFC in terms of
rheological properties and reaction mechanism. It is, however, slightly less retarding. SMFC is
limited to use in temperatures less than 185F (85C) because of its chemical stability.

Polystyrene sulfonates are effective dispersants but are rarely used, principally because of cost
in comparison to the SNFC. Contrary to the SNFC and SMFC, the polystyrene sulfonates appear
to have better rheological properties at higher molecular weights. This is probably because the

48
high-molecular-weight material shows more of a loop-type adsorption, thus placing more
negative charge on the surface and also producing some steric hindrance (Figure 2.32).195

Hydroxycarboxylic acids may be used as the primary dispersant in freshwater slurries at higher
temperatures (BHCT 200F). This is typically advantageous with cements that have a high free
alkali (> 0.75%) content to offset their retarding properties. Hydroxycarboxylic acids, such as
citric acid, are also used to disperse salt- and seawater cement slurries. The concentration of use
is limited by the temperature and thickening time desired, although concentrations of 0.5 to 1.0
% BWOC are usually sufficient.

2.11.6 Fluid-Loss Control Additives

Fluid-loss control additives are used to maintain a consistent fluid volume within a cement slurry
to ensure that the slurry performance properties remain within an acceptable range. Thus far, we
have discussed additives that address individual parameters such as
thickening time (accelerators and retarders)
density (lightweight additives and weighting agents)
viscosity (dispersants)
compressive strength (accelerators).

All of these parameters are critical to a successful cementing operation, and the variability of
each of these parameters is dependent upon the water content of the slurry. For example,
If the water content is greater than intended, the following will normally occur:
o Increase in thickening time, fluid loss, free fluid, sedimentation, permeability, and
porosity
o Decrease in density, viscosity, and compressive strength
If the water content is less than intended, the opposite will normally occur. The
magnitude of change is directly related to amount of fluid lost from the slurry.

Because predictability of performance is typically the most important parameter in a cementing

operation, considerable attention has been paid to mechanical control of slurry density during the
mixing of a slurry to assure reproducibility. Of equal importance is the slurry density during
displacement, which is directly related to fluid-loss control.

Cement slurries are colloidal suspensions consisting of distinct solid and liquid phases. During
the cementing operation, several opportunities exist for the fluid phase to separate from the
cement slurry. This can occur when the slurry is passing through small orifices or ports, and
within the annulus. When the slurry is passing through orifices, the fluid phase can be
accelerated, resulting in particle bridging. In the annulus, fluid can be displaced from the slurry
while it is passing though constricted areas, or lost to the formation, resulting in an increase in
the equivalent circulating density (ECD) which can lead to formation fracture (lost circulation)
or flash set (dehydration). After placement, the fluid phase will filter to permeable formations,
reducing the slurry volume and effective hydrostatic pressure and creating the potential for the
migration of formation fluid into and through the cement column. Fluid-loss control additives
help prevent solids segregation during placement and help control the rate of fluid leakoff in the
static state.

Neat cement slurries normally exhibit an uncontrolled API fluid loss of at least 1500 cc/30
minutes. This value is excessive for most cementing operations where permeable formations are
encountered or where long columns of cement will be used. The amount of fluid-loss control
49
required for a particular operation varies widely, and is largely dependent upon the slurry
density, the water content, the formation properties, and annular clearance.

Several materials are effective as fluid-loss control additives. Materials currently in use can be
loosely categorized into two groups according to their solubility characteristics: water-insoluble
and water-soluble. With the exception of bentonite, the water-insoluble materials are polymer
resins. All of the water-insoluble materials function as permeability reducers. The water-soluble
materials are modified natural polymers, cellulosics, and vinylinic-based polymers. The
polymeric materials, whether water-insoluble or -soluble, are all synthetic (manmade) materials.
Cellulosics are commonly referred to as chemically modified natural polymers. The water-
soluble polymeric materials are predominately anionic. However, in the case of the vinyl
polymers, cationic groups can be included. The water-soluble polymers are of higher molecular
weight (Mw 30,000) than their low-molecular-weight cousins (e.g., lignosulfonate and
dispersants) and are classified as polyelectrolytes.

The action of fluid-loss control additives depends on their solubility. The water-insolubles
reduce the permeability of the filter cake developed. Their effectiveness is dependent upon the
degree of interaction with the cement and hydration reaction byproducts (CH). They physically
obstruct the pore throat of the cement matrix, reducing the relative permeability of the filter cake.
The polyelectrolytes are controlled by their hydration characteristics and their ionic character per
repeating (mer) unit. Polyelectrolyte solutions exhibit both chainlike properties and properties
caused by their electrolyte groups and the electric charges borne by these groups.196 As discussed
in Section 2.11, the ionic composition of the solution plays a significant role in the hydration
process (Figure 2.11). Hydration characteristics of polyelectrolytes are affected by the
electrolyte groups that extend from the polymer backbone as pendant chains. The pendant chain
orientation, ionic character, and charge density will affect solution and slurry viscosity, and the
polymers susceptibility to CH attack. The polyelectrolytes ionic character per mer unit will
control their interaction with the aluminate and ferrite phases, and C-S-H. As a result of these
interactions, the surface charge distribution of the initially formed hydration products is altered
(Figure 2.15), changing the particle orientation of the cement matrix from a single effective
body to several interconnected concentrated bodies (agglomerates) with a lower effective
permeability within the body. The relative size of the agglomerates (degree of concentration or
flocculation) is dependent upon the spatial arrangement of the polyelectrolytes pendant chains
(conformation) and ion character, and average molecular weight (chain length) of the polymer
itself. In general, as molecular weight is increased, the flocculation power of a polymer and its
viscosity contribution to the aqueous phase is increased. This combined effect, in conjunction
with control of the concentration used, controls the agglomerate density and buoyancy.
Optimization of the agglomerate density controls the rate of collapse of the agglomerate
(deposition of filter cake) and the permeability of the cake developed, resulting in a controlled
rate of fluid loss.

2.11.6.1 Water-Insoluble Materials

Bentonite is not typically the primary fluid-loss agent in slurries of normal density. In low-
density slurries where higher bentonite concentrations may be used, it may provide sufficient
fluid-loss control (400 to 700 cm3/30 minutes) for safe placement in noncritical well
applications. Bentonite allows fluid-loss control by reduction of filter cake permeability due to
pore throat bridging. Fluid loss rates can be erratic because of the concentration used at a given
density, variations in platelet disassociation due to shear, and stacking arrangement in the filter

50
cake. The addition of lignosulfonates usually improves fluid-loss control by dispersing non-
prehydrated bentonite. Dispersion will increase platelet disassociation, reduce flocculation, and
improve stacking arrangement in the filter cake. The fluid-loss rate of prehydrated bentonite
slurries is significantly higher than non-prehydrated slurries.

Microsilica imparts a degree of fluid-loss control to cement slurries because of its small particle
size, Dp 5 microns. The small particles reduce the pore throat volume within the cement matrix
through a tighter packing arrangement, reducing filter-cake permeability. Microsilica is
particularly effective when used in conjunction with polyelectrolyte fluid-loss control additives.
The combined effect allows extremely low fluid-loss values to be obtained with a minimum
addition of fluid-loss control additives. It also allows for the use of higher dispersant
concentrations without the normal loss in slurry stability. The combined use of microsilica, fluid-
loss control additives, and dispersant is useful when both low fluid loss and viscosity are desired,
such as in gas-tight cements.

Polyvinyl alcohol (PVA) is a white to cream-colored powder with a density range of 1.27 to
1.31 g/cm3. It is a water-soluble polymer derived from polyvinyl acetate, and is chemically
reactive with acids and alkalies. It is not categorized as a water-soluble polymer because it loses
solubility in alkaline environments such as the aqueous phase of a cement slurry. PVA is derived
by alkaline hydrolysis of polyvinyl acetate. Maximum water solubility occurs at approximately
88% hydrolysis, and at a hydrolysis of 87 to 89%, polymers are readily soluble in both cold and
hot water.197,198 As the degree of hydrolysis increases, hydrogen bonding increases and water
solubility decreases, and as hydrolysis reaches 95%, water solubility is lost as the polymer
becomes crystalline.

PVA is not chemically stable in cement slurries. Because of PVAs narrow solubility range and
the alkaline nature of cement slurries (Section 2.10.2), hydrolysis occurs rapidly and solubility is
lost, resulting in a swollen pliable particle with a high degree of intermolecular hydrogen
bonding and cross-linking. The insoluble particle is effective in blocking the matrix permeability
of the filter cake if sufficient concentration is used. As with all reactive permeability reducers,
performance is characterized by a threshold concentration effect (Figure 2.33). Because PVA
can crosslink with the hydrated cement, bonding and set cement permeability are normally
improved. However, compressive-strength development is generally reduced. Thickening time
can be erratic because of variations in the rate of hydrolysis. Thickening-time reproducibility and
fluid-loss control can be improved with the addition of
CaCl2 to accelerate crosslinking
ethoxylated compounds to inhibit crosslinking199
retarders to inhibit the aluminate, ferrite, and C-S-H phases

SNFC and SMFC dispersants should be avoided unless used in conjunction with ethoxylated
compounds. Because SNFC and SMFC contain formaldehyde, they are highly reactive with PVA
and will undergo immediate polymerization (condensation) on contact in the aqueous phase,
resulting in extreme slurry gelation.

PVA is generally used at BHCT 130F (54C) at concentrations of 0.75 to 1.5 % BWOC,
depending on slurry density and composition.200 Fluid-loss rates of less than 50 cc/30 minutes
are common, and linear fluid-loss rates are not possible because of the threshold concentration
effect. Increased BHCT applications may be achieved with appropriate compounds, although
PVAs use is generally limited to less than 200F (93C).199 The main advantages of PVA are
51
that it can be used to obtain extremely low fluid-loss values and it is compatible with CaCl2. The
disadvantages of PVA are its effect on compressive strength and rheology (Table 2.24).

Synthetic latex is an oil-in-water emulsion system consisting of a dispersed phase of a water-

insoluble elastomer, surfactants, and a water exterior phase.196-198,201 These emulsions are
characterized by a milky white appearance. Their density is typically 1 g/cm3. The elastomer
system in common use is styrene-butadiene rubber (SBR), a copolymer of styrene-butadiene
with an active solids content of approximately 50%.201,202 The surfactant system concentration is
normally at least 3% of the product package, and usually contains at least one ethoxylate.202 The
surfactants help prevent separation and coagulation of the copolymer particles. The copolymer
particle sizes range from 0.05 to 0.15 microns, and exhibit colloidal properties.

The surfactant system plays a key role in the use of latex in well-cementing applications. In
cement slurries, the emulsion system readily disperses and exhibits time, shear, and temperature-
dependent stability. The emulsion stability can be improved with additional surfactant and,
depending on the surfactant type and concentration, the emulsion stability may be controlled to
above 300F (149C) BHCT. The surfactant system also acts as a dispersant in the cement slurry,
resulting in low slurry viscosity. Control of emulsion stability is critical to slurry performance
because the emulsions rate of inversion controls slurry viscosity and thickening time. Inversion
of the emulsion system results in an almost instantaneous conversion to a rubberized mass (set)
that is reported as the pumping time for the slurry.

Latex is typically used at a concentration of 0.8 gal/sk (~3.5 % BWOC dry weight equivalent)
to obtain a fluid loss of less than 100 cc/30 min. Fluid-loss values of less than 20 cc/30 min are
possible at a 1.5 gal/sk concentration in non-silica slurries and 2 to 3 gal/sk concentrations with
35 % BWOC silica slurries (Fig. 2.34).203,204 Fluid loss is controlled in the cement slurry by
particle plugging. The colloidal property of SBR and the high active solids content allows the
polymer particles to coagulate in the pore throat of the filter-cake matrix, resulting in very low
permeability. The combined effect of low permeability, dispersion, and rapid inversion provides
gas migration-control properties. Because of the high concentration of use, water content of the
slurry must be considered. If water content is insufficient, compressive strength development
will be significantly delayed and reduced at lower temperatures. This can be addressed by the
adding weighting agents to increase the mixing water content to at least 26 % BWOC. An
advantage of latex cements is their improved permeability, bond, tensile, and flexural
strengths.202 The main advantage of latex is its ability to impart gas-tight properties to the
cement.203,204 The main disadvantages of latex are the high concentrations required and its effect
on compressive strength.

2.11.6.2 Water-Soluble Materials

Derivatized Cellulose. Chemically, cellulose is a straight-chain polymer of D-glucose units
linked by -(1,4)-glucoside bonds.201 It may be derivatized to form esters and ethers, of which
most are water-insoluble. Upon sufficient substitution, the carboxymethyl (CM) and
hydroxyethyl (HE) ethers become readily soluble in water.197 Generally, water solubility
increases with increased substitution. In their unoxidized form, CM and HE ethers are of high
molecular weight (~500,000) and are strong viscosifiers. The materials used in cementing have
been oxidized to reduce their molecular weight to the 40,000- to 60,000-range.

52
Two forms of derivatized cellulose have been found useful in well-cementing applications. They
are the single-derivatized hydroxyethyl cellulose205 (HEC) and twice-derivatized carboxymethyl
hydroxyethyl cellulose (CMHEC).206,207 The usefulness of the two materials is largely dependent
upon their retardational character (Section 2.11.2) and thermal stability limits.

HEC is commonly used at temperatures up to 180F (82C) for fluid-loss control, and may be
used at temperatures up to 230F (110C) BHCT, depending on the co-additives used and slurry
viscosity limitations. Above 230F (110C), HEC is not thermally stable because of oxidative
attack of the ether linkage in the polymer backbone, resulting in a reduced molecular weight. As
molecular weight is reduced, HEC loses its effectiveness as a fluid-loss control additive.

HEC is a nonionic polymer (Figure 2.35). However, in solution, HEC is electrochemically a

weak polybase that can bind a proton, yielding a polycation of low ionic character per mer unit
relative to aluminate, ferrite, and C-S-H phases. Consequently, HEC is not very effective at
altering the surface charge distribution of the initially formed hydration products. Its ionic
character and lower molecular weight reduces its effectiveness at altering particle orientation in
the cement slurry and increases its dependence on fluid viscosification to control fluid loss. As a
result, when HEC is used alone, slurries become quite viscous, limiting the use of HEC alone in
saltwater, seawater, and low-density slurry applications.

In most applications, HEC is used with an anionic co-additive to optimize performance. The
most commonly used co-additive is the anionic dispersant SNFC. This combination is
particularly beneficial in that it combines particle dispersion and aqueous phase viscosity to
produce acceptable slurry viscosity and improved fluid loss. Low-molecular-weight (~30,000)
nonionic polymers, such as polyvinyl pyrrolidone (PVP)208 and/or hydroxycarboxylic acids, may
also be used to improve fluid-loss control.

HEC is typically used at a concentration of 0.4 to 3.0 % BWOC at densities of 16.0 to 11.0 lb/gal
at temperatures ranging from 80 to 150F (27 to 66C) BHCT to achieve a fluid loss of less than
100 cc/30 min (Fig. 2.36). When used at a 1:1 ratio with SNFC, HEC is typically used at a
concentration of 0.6 to 1.25 % BWOC at densities of 16.0 to 14.8 lb/gal at temperatures ranging
from 100 to 180F (38 -82C) to achieve a fluid loss of less than 100 cc/30 min. A 1:1:0.5 ratio
of HEC:SNFC:PVP provides improved fluid loss and lower slurry rheology at all temperatures
and is preferred at 175F BHCT and above. Due to potential retardation problems, the
hydrocarboxylic acids are generally limited to the upper temperature range of HEC and to
concentrations of approximately 0.15 % BWOC, which is sufficient to improve fluid loss.

CMHEC is commonly used at temperatures up to 300F for fluid-loss control (Figure 2.37), and
may be used at temperatures up to 350F, depending on the degree of substitution, the co-
additives used, and slurry viscosity limitations. Since CMHEC is twice-derivatized, it is more
thermally stable than HEC, and is not as susceptible to oxidative attack. The improved thermal
stability is directly related to the degree of carboxymethylation and hydroxyethylation achieved
by increasing the rigidity of the polymer backbone and sterically hindering oxidative attack.

CMHEC is a net anionic polymer that contains both anionic CM and nonionic HE electrolyte
groups (Figure 2.35). A weak polybase has little affect on surface charge density or particle
orientation. However, in solution, the CM electrolyte group is a weak polyacid that will
dissociate into polyanions and positively charged protons (H+), yielding strong ionic character
per mer unit. Consequently, CMHEC is effective at altering the surface charge density, particle
53
orientation, and permeability of the agglomerate body. As a result, improved fluid loss is
obtained at a lower slurry viscosity (Figures 2.38 and 2.39).

The effectiveness of CMHEC may also be improved by adding anionic co-additives. The
common co-additives are SNFC, SMFC, and PVP. Because of CMHECs increased temperature
range of application, hydroxycarboxlic acids (HCA) may also be used. The antioxidant and
sequestrant properties of hydroxycarboxylic acids are particularly well suited to improving the
stability of CMHEC at temperatures greater than 300F BHCT (149C).209 Compared to SNFC,
SMFC, and PVP, the hydroxycarboxlic acids are much more effective. Of the derivatized
celluloses, CMHEC is more effective in salt-containing slurries. If used as a co-additive with
other fluid-loss additives, it will act as a fluid-loss enhancer.

CMHEC is used at a concentration of 0.15 to 0.3 % BWOC at temperatures of 150 to 200F (65
to 93C), or 0.5 to 1 % BWOC at temperatures of 180 to 300F (82 to 149C), BHCT to achieve
100 cc/30 min. As a fluid-loss enhancer, CMHEC is typically used at a 1:1 ratio. If used in
conjunction with HCA, a minimum of 0.15 % BWOC HCA is used at a range of 5.67:1 to 1:1 of
CMHEC:HCA, in which the higher-molecular-weight HCAs are more effective. The
CMHEC:HCA combination is particularly effective with high HE- and low CM-substituted
CMHEC because of the reduced acid content (low CM). The 1:1 CMHEC:SNFC is commonly
used at temperatures less than 200F BHCT to provide retardation and fluid-loss control.

2.11.6.3 Synthetic Polymers

Synthetic polymers are a class of polyelectolytes that has recently found acceptance in well
cementing. Early work in this area indicated that the homo-polymers (single repeating mer unit)
and co-polymers (two or more repeating mer units) developed from existing monomers were not
chemically stable in the alkaline environment of cement slurries. In the late 1960s and early
1970s, several new water-soluble monomers became available commercially. Since the 1970s, a
significant amount of work has been done to develop synthetic co-polymers for use in cement
slurries using these monomers. Most of this work has centered on co-polymers of acrylamide
and/or acrylamide derivatives and their salts; however, several non-acrylamide-based monomers
have also been reviewed. The following list contains the acid form of the monomers of current
importance for use in co-polymers for well-cementing applications and their commonly used
acronyms.
acrylamide (AM)
acrylic acid (AA) or salt (SAA)
itaconic acid (IA)
acrylonitrile (AN)
2-acrylamido-2-methylpropane sulfonic acid (AMPS) or salt (SAMPS)
vinylsulfonic acid (VS) or salt (SVS)
styrenesulfonic acid (SS) or salt (SSS)
N-vinylpyrrolidone (NVP)
N,N, di-methylacrlylamide (NNDMAM)
N-vinyl-N-methylacetamide (NVNMA)
methacrylamidopropyltrimethylammonium chloride (MAPTAC)

Several combinations of these monomers have been used to produce co-polymers that impart
fluid-loss control to cement slurries. The goal of this work has been to produce polymers that

54
function as fluid-loss-control additives over a wide temperature range without the limitations of
the derivatized celluloses.

In the early work, the sodium salt of the random co-polymer product of AM-AA was found to be
effective for fluid loss control.210 However, the use of AM-AA was limited by the unpredictable
nature of the neutralization step and the uncontrolled alkaline hydrolysis of the AM fraction
when present in a cement slurry. Because of the alkaline hydrolysis, the anionic character of the
polymer and its reactivity with the aluminate and ferrite phases is increased, resulting in
unpredictable slurry retardation.211 The problem was addressed with some success by co-
polymerization of AM with monomers that were less susceptible to alkaline hydrolysis. The
polymerization of AM monomer with the nonionic polymer (PVP)212 to form a triblock polymer
and the controlled hydrolysis of a co-polymer of AM and the anionic monomer (AMPS)213
provided improved fluid-loss control. However, both copolymers were still susceptible to
alkaline hydrolysis, limiting their temperature range of use to less than that of the derivatized
celluloses. As a result, work continued, leading to the development of the co-, ter-, and tetra-
polymers that are in use today.

Initially, work continued with the development of co-polymers. Because of the susceptibility of
AM to alkaline hydrolysis, alternate monomers were used. Random co-polymers of an acid
monomer and a nonionic monomer such as SSS-NVP214,215, AA-NVP216, and AMPS-
NNDMA217-219 were preferred. Concurrently, ter- and tetra-polymers were being developed that
provided further improvements. The ter- and tetra-polymers are also based on acid-nonionic
monomer combinations that include, SVS-AM-NVNMA220, AMPS-AM-NVNMA,220 SVS-AA-
NVNMA,221 AMPS-AA-Styrene,222 AMPS-AM-IA,223 and AMPS-AM-NVP-AA.224,225 Further
developments included polyampholyte ter-polymers of SSS-AM-MAPTAC226 and vinyl-grafted
wattle-tannin polymers containing AMPS or AMPS-AM or AMPS-NNDMA or AMPS-AN-
NNDMA.219,227

Of the polymers listed, the co-polymers of AMPS-NNDMA, the ter-polymer AMPS-AM-

NVNMA, the tetra-polymer of AMPS-AM-NVP-AA (Figure 2.40), and the vinyl-grafted wattle
tannins have the widest acceptance. The AMPS-AA-NNDMA and AMPS-AM-NVP-AA have
the unique characteristic of not extending thickening time or retarding the rate of strength
development. The AMPS-AM-NVNMA polymer undergoes alkaline hydrolysis of NVNMA
repeating unit beginning at approximately 150F (66C) BHCT, resulting in increased
retardation and a reduction in slurry viscosity. However, the fluid-loss property is improved. The
vinyl-grafted wattle tannins retard at all temperatures and have a strong dispersing effect. The
dispersing effect is particularly useful in cements with a high alkali content or in weighted
slurries in which water content is limited. All of these materials are effective to over 300F
(149C) BHCT, and their temperature range can be extended with HCA. In general, fluid-loss
control is improved by the use of ethoxylated surfactants and common dispersants, although in
the case of dispersants, the concentration of use may be limited (Figure 2.41).

Poly Vinyl Pyrrolidone (PVP) is a nonionic polymer that is typically used as a fluid-loss
enhancer in conjunction with SNFC to improve the performance of other polymers. When used
alone, PVP is not very effective as a fluid-loss-control additive. However, when used in
conjunction with SNFC, the fluid loss is improved through improved particle orientation.228
PVP/SNFC is particularly advantageous for dispersion and fluid-loss control in densified
cements. The use of PVP/SNFC with HEC208 or CMHEC significantly improves fluid-loss
control.

55
Surfactants are surface-active agents that modify the interfacial tension between two liquids or
between a liquid and a solid. Surfactants play a significant role in the interaction of colloidal
particles. Low-molecular-weight polymers such as SNFC, SMFC, and lignosulfonate are
surfactants. Polymers commonly used as fluid-loss-control additives are macromolecules that fall
into the lower end of the 1-milli-micron to 1-micron size range for colloidal particles. As a result,
the choice of surfactant can significantly affect the fluid-loss-control additive itself and its
interaction with cement particles.

Surfactants can be used to accelerate or retard the solubility or wettability of polymers, to form
branches between polymers or the cement particles and polymers, and to sterically hinder
(stabilize) the electrolyte groups on the polymer chain. In the case of synthetic latex, surfactants
retard the wetting characteristics of the SBR particles, increasing the fluid life of a latex cement
slurry.203,204 Surfactants sterically hinder the reactive electrolyte groups in PVA, improving fluid
loss and rheology.199 Proper surfactant selection can improve the fluid-loss performance of most
polymers as well as decrease shrinkage and improve interfacial sealing.229

2.11.7 Lost-Circulation Additives

Cement slurries can be lost to the formation and not circulated back to the surface during
completion of a wellbore. 230-232 This is defined as lost circulation. It should not be confused with
the volume decrease resulting from fluid-loss filtration (Section 2.11.6). Lost circulation
typically occurs in three basic formation types:
unconsolidated or highly permeable - particles of a cement slurry can enter an Formatted: Indent: Left: 0.5", Bulleted +
Level: 1 + Aligned at: 0" + Tab after: 0.25" +
unconsolidated or highly permeable formation only if the permeability is greater than Indent at: 0.25"
100 darcies.
fractured, induced or natural - induced fractures occur in highly incompetent zones Formatted: Indent: Left: 0.5", Bulleted +
Level: 1 + Aligned at: 0" + Tab after: 0.25" +
(i.e., shale) that break down at relatively low hydrostatic pressures. Natural fractures Indent at: 0.25"
can be encountered anywhere.
cavernous or vuggy - these caverns or holes are usually formed by erosion of the Formatted: Indent: Left: 0.5", Bulleted +
Level: 1 + Aligned at: 0" + Tab after: 0.25" +
formation caused by the action of subsurface waters, and are discovered Indent at: 0.25"
unexpectedly.

In many cases, lost circulation occurs during drilling as evident by the loss of drilling fluids, and
actions can be taken at that time to combat the problem. At other times, difficulties may be
encountered during drilling, indicating potential lost-circulation problems and measures can be
taken to prevent their occurrence during cementing. Typically, combating lost circulation
involves two steps: (1) reducing slurry density and (2) adding a bridging or plugging material.233-
235
Additives for preventing lost circulation can be separated into three basic groups:
bridging materials Formatted: Indent: Left: 0.5", Bulleted +
Level: 1 + Aligned at: 0" + Tab after: 0.25" +
rapid-setting or thixotropic cements Indent at: 0.25"
lightweight cementing systems

Bridging materials physically bridge over and/or plug the lost-circulation zone and typically
come in fibrous, flake or granular forms. Most bridging materials are considered to be
chemically inert with respect to cement hydration.
Fibrous materials are generally used for controlling lost circulation in highly Formatted: Indent: Left: 0.5", Bulleted +
Level: 1 + Aligned at: 0" + Tab after: 0.25" +
permeable formations. In effect, they stop flow by forming a mat over the surface. Indent at: 0.25"
They can also be used to channel into and effectively plug small fractures. The

56
 concentration of fibers is usually low0.125 to 0.25 lb/skbecause of mixing and
equipment issues.
The most common flake material is cellophane. Cellophane flakes are typically 0.02
to 0.06 mm in thickness and 0.94 cm to 1.87 in length. They act by forming mats or
bridges over very narrow fractures. Concentration range of cellophane is usually
0.125 to 0.5 lb/sk.
Granular materials that are most frequently used include gilsonite, perlite, and coal.
These coarse particles are typically used for large fractures and cavernous or vuggy
lost-circulation formations. As the cement slurry enters the formation, these large
granular particles, in principle, become trapped and block off the opening. They also
cause restrictions where the opening is smaller than the largest granules present.
Concentrations vary according to the material used and are typically, 5 to 50 lb/sk for
gilsonite, 0.5 to 1.0 ft3/sk for perlite, and 1 to 10 lb/sk for coal. These granular
materials also act as lightweight materials (Section 2.11.3.1) in reducing the
hydrostatic pressure and helping prevent formation breakdown in weak zones. Other
granular materials that are less frequently used include chopped rubber, ground
walnut or pecan shells, rice husks, coarse bentonite, corn cobs, and straw. In cases
where the slurry loss is very rapid, any bridging material available on site may be
used.

In large cavernous or vuggy formations where bridging materials are not effective, rapid-setting
and thixotropic cements are the preferred means for controlling lost-circulation. These cements
are usually designed to set up in the lost-circulation zone, plugging it off.

Rapid-setting cements include both quick- and flash-setting formulations. These cements
generally produce thin slurries, but have very short setting times. The quick-setting cements will
set up while being displaced or shortly after entering the lost-circulation zone, whereas the flash-
setting cements form semi-solid materials when mixed with water or water-based drilling fluids.

Thixotropic cements have a low viscosity during mixing and placement, but when they enter the
formation and are no longer subjected to shear, they gel and become self-supporting. Thixotropic
formulations include gypsum cement, gypsum Portland cement, aluminum sulfate/iron (II)
sulfate, clay-based systems, and crosslinked polymer systems. Gypsum cement and gypsum
Portland cement systems are based on hemihydrate-forming gypsum (Section 2.9.3); the gypsum
further reacts with the aluminate and ferrite phases to form ettringite (Figure 2.9). It is the
increase in the amount of ettringite formed that causes the thixotropic behavior. The aluminum
sulfate/ iron (II) sulfate system works on the same principle. Clay-based systems, typically
bentonite, rely on the development of gel strength due to the swelling of the clay to impart some
thixotropy to the system. Crosslinked polymer systems consist of a water-soluble polymer such
as hydroxyethylcellulose (HEC), carboxymethylhydroxyethylcellulose (CMHEC), and polyvinyl
alcohol (PVA), and a crosslinking agent that includes titanium or chromium chelates.235 The
thixotropic behavior depends on the optimum polymer/crosslinker combination, relative
concentrations, and temperature of the well. The properties of the different thixotropic systems
vary considerably, and the selection of a slurry system should be based on the specific problem
of the well.

Lightweight additives are used primarily as a preventive measure when drilling problems
indicate lost circulation and/or in characteristically weak formations. The lower-density slurries
reduce the amount of hydrostatic pressure in the well and minimize the risk of exceeding the

57
threshold of the formation fracture pressure. Lightweight additives and extended systems are
discussed in Sections 2.9.1 and 2.11.3.

Combining bridging materials with rapid-setting or lightweight systems is often a more effective
solution for lost circulation. The choice of a system and bridging materials will depend on the
type of formation, the size of the lost-circulation zone, the fracture pressure gradient, and the
downhole temperatures and pressure, as well as economics.

2.11.8 Strength Retrogression Inhibitors

Strength retrogression, a phenomenon that occurs naturally with all Portland cements at
temperatures of 230 to 248F (110 to 120C), is usually accompanied by a loss in
impermeability, and is caused by the formation of large crystals of -dicalcium silicate hydrate
(-C2SH). Silica flour or silica sand is commonly used to prevent strength retrogression by
modifying the hydration chemistry, and it can be used with all classes of Portland cement. (See
Section 2.10.3). The addition of 30 to 40% silica is usually adequate to produce a set cement
with low permeability (<0.1 millidarcy) (Table 2.25) that overcomes the problems of strength
retrogression, though additions can range from 30 to 100%.119,236 Silica flour is mined and
processed to give a powder with a particle size of < 75m ( 200 mesh) for maximum reactivity
in normal and lightweight cements. It has a water requirement of 40% by weight of silica flour
(4.8 gal of water/100 lbm of silica flour). Silica sand is a coarsely ground silica with particle
sizes of 75 to 210m (-50 to +150 mesh). Silica sand has no water requirements and is designed
for densified cements above 17.2 lb/gal. It is not recommended for lower-density slurries
because the larger particles tend to settle out.

2.11.9 Free-Water Control

In well-cementing applications, the maintenance of a consistent column of cement is critical to
assure proper zonal isolation. Because of rheological demands and the need for silica or
weighting agents in some applications, consistency is not always possible with conventional
materials. An additional additive is often required to address the potential problem of particle
sedimentation. This group of additives is known as free-water control additives.

Sodium silicate may be used to control free water in normal- to low-density cement slurries.
Sodium silicate reacts with CH to form additional C-S-H gel on the cement particle surface,
resulting in localized particle enlargement (Section 2.11.3.3). Typically, 0.15 to 0.5 % BWOC is
sufficient to provide free-fluid control.

Biopolymers thin a cement slurry at high shear rates and viscosify the slurry at low shear rates.
As a result, slurries more readily go into turbulent or upper laminar flow, yet have sufficient low
shear to prevent sedimentation. Xanthan gum and Welan gum both provide these characteristics,
and are typically used at an active concentration of approximately 0.2 % BWOC.

Synthetic polymers of high molecular weight that are resistant to alkaline hydrolysis are
effective as free-fluid-control additives at temperatures where sodium silicate and biopolymers
are not effective. These are ridged polymers that maintain their viscosity or increase in viscosity
with increasing temperature. They are typically used at an active concentration of 0.1 to 0.2 %
BWOC.

58
2.11.10 Expansive Cements
Expansive cements are used primarily for improving the bond between the cement and the pipe
and the cement and the annulus for effective zonal isolation. Good zonal isolation is essential to
prevent loss of production, control gas migration, provide protection from corrosive formation
waters, reduce water production, and improve confinement of stimulation treatments. Poor
bonding of cement to pipe and/or annulus most often results from a combination of effects from
a variety of factors. The root causes are usually associated with drilling-fluid properties and
displacement mechanics, casing expansion and contraction caused by thermal stresses or internal
pressures, fluid loss from the cement, and hydration volume reduction during setting of cement.
The resultant effect of poor bonding is the formation of microannuli or small gaps at the
cement/casing or cement/formation interface. Expansive cements expand slightly after the
cement has set to fill any void spaces. Because of the restraints imposed by the casing and
formation, any additional expansion will occupy space provided by the cements internal
porosity. The two principal types of expansive additive or cement are post-set crystalline growth
or chemical expansion and in-situ gas generation.

Crystalline-growth additives. The additives used to expand cement through post-set crystalline
growth are either Aft-forming or polyvalent metal oxides that exhibit delayed hydration or
cements containing high concentration of salt. The expansion mechanism is the growth of the
crystals within the solid cement matrix. Since the crystals have a greater bulk volume than the
original solids from which they form, they cause a wedging action due to the internal pressure of
crystalline growth, forcing the solid matrix apart. Crystal growth expansion is unilateral in that
restraint in one direction does not increase expansion in other directions. The amount of
expansion is dependent on the amount of additive, curing time and temperature, and, in some
cases, cement-slurry composition.

Several Aft-forming expanding cements and cement additives are available. The expanding
cements are discussed in detail in Section 2.9. Expanding additives are typically calcium sulfates
such as C S and C S H0.5 that have a defined particle size and solubility. These are used with
either a Portland cement having a high aluminate (substitutedC3A) content (> 5%) or with an
additional source of aluminum such as Al2(SO4)3. The calcium sulfate-Portland cement system is
similar to that of the Type S cement, but it can be prepared as needed. The amount of expansion
is dependent on the fineness and aluminate content of the cement used. The AFt formed is
similar to that produced during cement hydration (Figure 2.9), but it is designed to form in
addition to the normal AFt and after the cement has gained some compressive strength. Aft-
expanding cement systems are generally thixotropic, although these properties can be minimized
by adding dispersant. These expanding cement systems are limited to the temperature stability of
the Aft phase which is approximately 158F (70C). 237

Polyvalent metal oxides used as expansive additives are typically calcined (dead-burnt) at
temperatures of 2192F (1200C) to decrease their solubility so that they do not react before the
cement sets.238 The two products most commonly used are CaO and MgO (periclase). A dead-
burnt CaO expanding cement has been produced commercially (Section 2.9). The basic
hydration reaction is:
RO + H2O R(OH)2
As with the Aft-expanding systems, the hydrated material occupies more space than the original
material. The hydration products of the polyvalent metal oxides are also thought to have an
additional layered water, giving R(OH)2:nH2O and further increasing their volume. Polyvalent

59
metal oxides can be used at temperatures of 130 to 550F (54 to 28C) and where the reactivity
can be controlled largely by the degree of fineness.

Cement slurries containing high concentrations of salt (NaCl, KCl or CaCl2) are also known to
cause expansion.239 Expansion is caused by the crystal growth of calcium chloroaluminate
hydrate (3CaO.Al2O3.CaCl2.xH2O) from reaction of the chloride ions with the aluminate phase in
cement. The temperature limitation for calcium chloroaluminate hydrate is believed to be
approximately 125F (51C), although salts are reportedly effective as expanding additives up to
400F (204C), depending on the system. Salt also improves bonding by preventing dissolution
of the salt formation.

In-situ gas generation. Expansion from in-situ gas-generating additives occurs before set, while
the cement is still in the plastic state. The most common in-situ gas-generating additive is
aluminum powder, although zinc, iron, and magnesium are possible alternatives. The expansion
is caused when the additive reacts with alkali and water present in the cement aqueous phase
(Section 2.9) to produce micro-size bubbles of H2 gas. Expansive forces that are a direct function
of the gas generated compensate for any volume losses caused by hydration volume reduction or
fluid loss and increase the pressure of the cement against the pipe and formation. In-situ gas-
generating additives can be used at temperatures of 60 to 400F (16 to 204C). Because of the
compressibility of the gas, the amount required is more dependent on the hydrostatic pressure of
the slurry than on the downhole temperature. Concentrations generally range from 0.15 to 0.6 %,
although they can be higher. Because the reaction is a surface reaction, the reactivity of the in-
situ gas-generating additive depends on the fineness of the particles. Intensifiers can be added to
increase reaction at lower temperatures. The one disadvantage of the in-situ gas-generating
additive is that the H2 gas it produces is highly flammable. Care should be taken when pumping a
slurry containing the additives and necessary precautions should be taken if it is to be circulated
back to the surface.

2.11.11 Miscellaneous Additives

Several additives do not fit in any of the preceding categories. These additives can be used
frequently (as with antifoam additives) or infrequently for specialized applications (as with mud
decontaminants, radioactive tracers, dyes, and fibers).

2.11.11.1 Antifoam Additives

Antifoam additives are frequently used to decrease foaming and minimize air entrainment that
arises during mixing. Foaming is a secondary effect caused by a number of additives. Excessive
foaming can result in:
an underestimation of the density downhole Formatted: Indent: Left: 1", Bulleted + Level:
1 + Aligned at: 0" + Tab after: 0.25" + Indent
cavitation in the mixing system at: 0.25", Tab stops: 1.25", List tab + Not at
0.25"
Slurry density is usually measured with a densitometer during mixing to proportion the solids
and water to obtain the desired density. When a slurry foams, the entrapped air is also included
in the density measurement and, because air compresses under pressure, the actual density
downhole becomes greater than that measured on the surface. Severe foaming can cause
cavitation of the pumps and ultimately lead to loss in hydrostatic pressure.

Antifoam additives, in general, modify the surface tension and/or dispersion of solids in the
slurry to prevent foaming or break up foam. The concentration of foaming additive required to be
60
effective is very small, typically less than 0.1% BWOW. Antifoam additives consist primarily of
polyglycol ethers or silicones or a mixture of both, and may also include surfactants.

Polyproylene glycol is the most common polyglycol ether used, and is preferred for its lower
cost. It is effective in most situations, although typically it must be added before mixing. In some
cases, it can interact with other additives and increase foaming.

The silicone antifoam additives are a suspension of very fine silica particles dispersed in a
silicone base and can also exist as an oil-in-water emulsion. They can be used both before and
during mixing, and are highly effective as antifoam additives.

2.11.11.2 Mud Decontaminant Additives

Paraformaldehyde or a blend of paraformaldehyde and sodium chromate is sometimes used to
minimize the cement retarding effects of various drilling-mud chemicals in the event a cement
slurry is contaminated by intermixing with the drilling fluids.240 A mud decontaminant consisting
of a 60:40 mixture of paraformaldehyde and sodium chromate neutralizes certain mud-treating
chemicals. It is effective against tannins, lignins, starch, cellulose, lignosulfonate, ferrochrome
lignosulfonate, chrome lignin, and chrome lignite. Mud decontaminants are used primarily in
openhole plugback jobs and liner jobs, for squeeze cementing, and for tailing out on primary
casing jobs.

2.11.11.3 Radioactive Tracers

Radioactive tracers are added to cement slurries to provide markers that can be detected by log-
ging devices. They were originally used to determine the location of fill-up or cement top and the
location and disposition of squeeze cement, although temperature surveys and cement-bond logs
now fulfill this function. Radioactive tracers are still occasionally used in remedial cementing to
locate the slurry after placement, if required, and for tracing lost circulation. The isotopes
commonly used downhole are 53I131 and 77Ir192, with half-lives of 7 and 84 days, respectively.
The 77Ir192 can also be tagged onto sand or glassy beads, primarily for use in lost circulation and
fracturing. Neither the tagged sand/glassy beads nor the 53I131 is recommended for squeeze
cementing. With the proper selection of tracer, the time required to get back into the hole for a
survey can be programmed. A baseline radiation log is usually run to determine the natural
radioactivity of the formation.

Radioisotopes are controlled and licensed by the U.S. Nuclear Regulatory Commission and
various state agencies, and cannot be used indiscriminately.

2.11.11.4 Dyes
Small amounts of indicator dye can be used to identify a cement of a specific API classification
or an additive blended in a cementing composition. When the dyes are used downhole, however,
dilution and mud contamination may dim and cloud the colors, rendering them ineffective. Table
2.26 lists some materials used as indicators.

Naturally occurring mineral oxides and/or synthetically produced color pigments may be
substituted for the dye indicator. To be effective, they should be finely ground to give more
uniform dispersion in the system. The amount required depends on the type of pigment and the
61
color desired. For example, a dose of pigment equal to 0.1 wt% of the cement may produce the
desired color, but more color may be needed to produce a deep contrast should contamination
occur downhole.

2.11.11.5 Fibers
Conventional Portland cement mixed at normal density has low ductility that makes it somewhat
brittle and thus, susceptible to post-cementing stresses. Synthetic fibrous materials are frequently
added to increase ductility and reduce the effects of shattering or partial destruction from
perforation, drill-collar stress, or other downhole forces.241 Fibrous materials transmit localized
stresses more evenly throughout the cement and thus improve the resistance to impact and
shattering. Nylon, with fiber lengths varying up to 1 in., has commonly been used because it is
resilient, and imparts high shear, impact, and tensile strength. The properties achieved with
nylon-fiber-reinforced cement on unsupported casing are illustrated in Figure 2.42. Particulated
rubber also improves a cements ductility and flexural strength, and is usually used in
concentrations up to 5% BWOC. More recently, aluminium silicate and/or fibrous calcium
silicates are reported to enhance the compressive, flexural, and tensile strengths.242 In general,
the aluminum silicate used is a high-reactivity metakaolin, and the calcium silicate is natural
wollastonite fiber. A typical concentration for each is 10% BWOC.243

2.11.12 Additive Summary

An overview of the most common cementing additives is given in Table 2.27. The table also
lists the primary uses and benefits of additives, and the cements they can be used with. The
primary effects of the cement admixtures on the physical properties of the cement, either as a
slurry or set cement, are illustrated in Table 2.28. This is a quick and ready reference; as such,
individual additives in a given category may not agree in total with the effects as illustrated. The
properties and effects are defined for individual additives and can be modified when additive
combinations are used.

2.12 Cementing Diagnostics

The methods of testing cement for downhole application are based on performance testing such
as listed in Section 2.11. Testing methods are usually performed in accordance with API
Specifications,25 though specifically designed and engineered equipment or tests are also used.
The choice of additives and testing criteria is dictated primarily by the specific parameters of the
well to be cemented. Performance testing has proven to be most effective in predicting how a
given slurry will behave under specific well conditions. There is no direct means of predicting
cement performance from the properties of cement as discussed in Section 2.5, and no technique
has been established or is likely to be established in the near future that would correlate cement
composition and cement/additive interaction with performance.

Performance testing, however, is not adequate in all situations, in particular, it is not effective for
troubleshooting downhole problems in which the integrity of the cement blend is in question.
Diagnostic analyses can be performed to evaluate the cement powder (Section 2.5), but no
definitive tests exist for chemically analyzing the composition of a cement once it has been
mixed with additives, as a dry blend, a slurry, or a set cement. The primary reason for this is the
low concentration of additives used. In the slurry or set cement the concentration of additives can
be even lower, as additives are consumed and/or possibly modified during the cement hydration
62
reaction. Analyzing downhole samples is often unreliable in that critical information such as
where the samples were obtained, and whether they were contaminated with drilling fluid or
formation waters during retrieval, cannot be confirmed. Many of the techniques used for
understanding the chemistry of cement are designed for laboratory-prepared specimens and
applications, and are not applicable to field samples. However, depending on the sample and the
concentration of additives, some qualitative analysis can sometimes be achieved.

Methods for determining the presence of cement:

XRD is the most appropriate method for analyzing dry cement powder (unreacted) or Formatted: Indent: Left: 0.5", Bulleted +
Level: 1 + Aligned at: 0" + Tab after: 0.25" +
cement that is not completely hydrated (Figure 2.3). It is also useful for determining Indent at: 0.25"
the hydrated phase composition(s) at temperatures above 302F (150C) (Figure
2.14). XRD is not advised for cements hydrated normally at temperatures below
302F (105C). In this case, only CH is typically present (Figure 2.14). If a
pozzolanic additive is used there would not even be a trace of the CH as it would be
consumed in formation of C-A-S-H which is amorphous. Below 122F (50C), some
AFt/AFm phases may be evident, depending on the amount of these phases and their
crystallinity, but the data cannot be considered absolutely conclusive.
Electron microscopy, coupled with microanalysis techniques, is one of the most
informative methods for cements at temperatures up to 302F (150C). It indicates
both the presence of C-S-H and the morphological characteristics. The CH and
AFt/AFm are usually also evident. Depending on the degree of preparation and the
microanalysis technique used, information on the actual chemical composition of the
phases can be obtained.105
Differential thermogravimetric analysis shows characteristic peaks that are associated
with the hydrated cement phases and with chemical changes that occur on heating the
sample. Water loss from C-S-H occurs at temperatures of 239 to 257F (115C
to125C), water loss from AFt at 275 to 284F (135 to140C), and CH decomposition
to CaO at 986 to 1022F (530 to 550C). AFm loses water in stages and produces
multiple peaks as temperatures increase; typically, in hydrated cement, AFm peaks at
approximately 365 to 392F (185 to 200C). The temperatures associated with the
peaks can vary with the techniques used, although the relative positions remain
almost the same.
FTIR-hydrated cement phases produce some specific absorption bands. However,
overlapping of the bands and the complexity involved reduce the usefulness of this
method.

Methods for determining the presence of additives at >15% BWOC

Electron microscopy coupled with microanalysis techniques can be used to determine Formatted: Indent: Left: 0.5", Bulleted +
Level: 1 + Aligned at: 0" + Tab after: 0.25" +
the presence of pozzolanic extenders such as fly ash or microspheres because of their Indent at: 0.25"
distinct spherical morphology. Diatomaceous earths can also be observed because of
their unusual morphology, as can most of the bridging agents used in lost-circulation
control. Microanalysis of individual particles or groups of particles is usually
effective in detecting additives such as microsilica, silica flour, and silica sand.
Chloride from salts can also be readily defined by microanalysis techniques.
Several different chemical analyses can be used to determine the presence of
chlorides. These chemical tests have been developed particularly for chlorides in
concrete. ASTM has a rapid chloride test for concrete.244
Methods for determining the presence of additives at <5%

63
 Chemical extraction and analysis of the aqueous phase is one possible route to Formatted: Indent: Left: 0.5", Bulleted +
Level: 1 + Aligned at: 0.25" + Tab after: 0.5"
determining the presence of additives used in small concentrations. In principal, + Indent at: 0.5"
extraction techniques give up to 70% recovery, although 50% or less is more typical,
because the additive is usually bound with the reaction products.245 The technique
used for analyzing the extracted sample depends on the type of additive in question
and the likelihood of interference from other additives. Because no defined extraction
procedures or solvent media for the various additives exist, extraction techniques
must be devised according to the chemistry of the additives involved. Techniques are
available for defining the additives if extracted, though the effects of interference
must be considered when combinations of additives are present. Similar studies have
been undertaken, primarily by Japanese researchers, in attempting to determine the
presence of additives in hardened concrete. Extraction techniques have included
room-temperature water or boiling K2CO3,245 ternary mixture of 75% methylethyl
ketone, 14% ethanol and 11% water,246 Na2CO3 extraction, boiling water extraction,
acid-chloroform extraction or nitric-acid digestion,247 and HCl.248,249 In many cases,
more than one extraction solvent was used. Methods of analyzing the extracted
samples have included:
high-pressure liquid chromatography (HPLC) for ionic surfactants250, SNFC, Formatted: Indent: Left: 1", Bulleted + Level:
1 + Aligned at: 0.25" + Tab after: 0.5" +
and retarders245,249 Indent at: 0.5"
ultraviolet spectroscopy (UV) for retarders, accelerators, and dispersants246,251
rapid chromatographic method for lignosulfonate248
pyrolysis-gas chromatography for organic admixtures252
infrared (IR) for retarders, accelerators, and water-repellent organic
admixtures
Titrimetric methods for lignosulfonates, hydroxycarboxylates, triethanolamine, and
glucose have been reported. 253 These methods are based on dissolution in H2SO4,
followed by the addition of NH4OH to remove Fe3+. The resulting organic matter is then
determined by titration with KMnO4.

In most cases, the concentration of additives remaining in the aqueous phase of oilwell cements
or slurries that can actually be extracted is so low that the additives cannot be detected by the
means outlined above.

Analysis of dry-blended samples is somewhat different from that of the slurry or set cement.
With sufficient quantities, performance testing is the most appropriate method for comparing the
actual blend with that designed. If quantities are not sufficient for performance testing, the blend
requires dissolution in an extracting solvent. This usually includes water and, inevitably, cement
hydration will occur, and some of the additive will be removed by the hydration products. After
the cement and additive are blended, separation of the additive from the dry sample is usually
impossible unless the additive has a significantly larger particle size or heavier density than that
of the cement.

The methods outlined require only small amounts of sample (milligrams to grams) and good
sampling procedures are required to give representative data.

References
1. Rondelet, J.: Lart de btir Paris, 1805.
2. Symposium on Use of Pozzolanic Materials in Mortars and Concretes, Special Tech. Pub.
No. 99, ASTM, Philadelphia, Pennsylvania (1949).
64
3. Parker J.: Cement or Tarras to be Used in Aquatic or Other Buildings and Stucco Work,
British Patent No. 2120 (1796).
4. Vicat L.J.: A Practical and Scientific Treatise on Calcareous Mortars and Cements, Artificial
and Natural, English Translation by Captain J.T. Smith, London (1837).
5. Aspdin, J.: An Improvement in the Modes of Producing Artificial Stone," British Patent No.
5022 (1824).
6. Hill J.R.: Improved Mode of Closing the Water Courses Encountered in Drilling Oil Wells,
US Patent No. 112,596 (1871).
1.7. American Petroleum Institute: Californias Oil, 12 (1948). Formatted: Indent: Left: 0", Numbered +
Level: 1 + Numbering Style: 1, 2, 3, + Start
2.8. Bye, G.C.: Portland Cement Composition, Production and Properties, The Institute of at: 7 + Alignment: Left + Aligned at: 0.01" +
Ceramics, Pergamon Press (1983). Tab after: 0.26" + Indent at: 0.26"
3.9. Ludwig, N.D.: Portland Cements and Their Application in the Oil Industry, Drill.
and Prod. Prac., API (1953) 183-209.
4.10. Jawed, I. and Skalny, J.: Alkalies in Cement: A
Review, Cem. Concr. Res., 7, 719-730 (1977).
5.11. Guinier, A. and Regourd, M.: Structure of
Portland Cement Minerals, 5th Int. Proc. Chem. of Cem., V 1, PT 1, 1-43 (1969).
6.12. Bogue, R.H.: Ind. Engnrg. Chem. (Analysis) 1, 192 (1929). Formatted: Indent: Left: 0", Numbered +
Level: 1 + Numbering Style: 1, 2, 3, + Start
7.13. Specification for Cements and Materials for Well Cementing, API Specification 10A, at: 7 + Alignment: Left + Aligned at: 0.01" +
Twenty Second Edition, API, Washington, D.C. (January 1995): Addendum 1, October 1999. Tab after: 0.26" + Indent at: 0.26", Tab
8.14. Taylor, H.F.W.: Modification of the Bogue Calculation, Adv. Cem. Res. (1989) 2, 73- stops: 0.25", Left + Not at 0.26"

77.
9.15. Campbell, D.H.: Microscopical Examination of Portland Cement and Clinker,
Construction Technology Laboratories, Skokie, Illinois 128 (1986).
10.16. Hofmnner, F. Microstructure of Portland Cement Clinker, Rheintaler Druckerei und
Verlag, Heerbrugg, Switzerland (1975).
11.17. Hjorth, L. and Laurn, K-G.: Belite in Portland Cement, Cem. Concr. Res. 1, 27-40
(1971)
12.18. Gutteridge, W.A.: On the Dissolution of the Interstitial Phases in Portland Cement,
Cem.Concr. Res. 9, 319- 324 (1979).
13.19. Neubauer, J., Pllmann, H., and Meyer H.W.: Quantitative X-ray Analysis of OPC
Clinker by Reitveld Refinement, Proceedings of the 10th International Congress on the
Chemistry of Cement, 3v007, 12pp (1997).
14.20. Campbell, D. H.: Innovations in Cement Manufacture, Chapter on Microscopy,
Portland Cement Association, in progress (2003).
15.21. Polkowski, G.: Correlation of Oilwell Cement Performance With Cement
Characteristics, paper SPE 16651 presented at the 1987 SPE Annual Technical Conference
and Exhibition, Dallas, September 27-30.
16.22. Fletcher, P. et al.: Predicting Quality and Performance of Oilfield Cements with
Artificial Neural Networks and FTIR Spectroscopy, paper SPE 28824 presented at the 1994
European Petroleum Conference, London, October 25-27.
17.23. Caveney, W.J. and Weigand, W.: Practical Oilwell Cement Microscopy, Proceedings
of the Seventh International Conference on Cement Microscopy, International Cement
Microscopy Association, Fort Worth, Texas, 37-55 (1983).
18.24. Annual Book of ASTM Standards, Section 4, Vol. 04.01, Cement, Lime, Gypsum,
ASTM, West Conshohocken, Pennsylvania, (2000).
19.25. Recommended Practice for Testing Well Cements, API Recommended Practice 10B,
Twenty-Second Edition, API, Washington DC (December 1997): Addendum 1, October
1999.

65
20.26. Halliburton Oil Well Cement Manual, Halliburton Co., Duncan, Oklahoma (1983).
21.27. Hansen, W. C.: Oil-Well Cements, Paper presented at the Intl. Symposium on the
Chemistry of Cement, London (1952).
22.28. Bensted J.: Oil-Well Cement StandardsAn Update, World Cement, 23, 38 (1992).
23.29. Cement Standards of the World, CEMBUREAU, European Cement Association
(1991)
24.30. ISOs 50th Anniversary: Friendship Among EqualsRecollections From ISOs First
Fifty Years, 96 p, (1997).
25.31. ISO in the New Century, 12 p, (1998): www.iso.ch
26.32. ISO Cement standards obtainable from API or global@ihs.com
27.33. Hewlett, P.C.: Leas Chemistry of Cement and Concrete, 4th Edition, London, Arnold
Publishers Ltd (1998).
28.34. Boncan, V.G. et al.: Method and Compositions for Use in Cementing, U.S. Patent
6,145,591 (2000).
29.35. Roy, D.M., Luke, K., and Diamond, S.: Characterization of Fly Ash and its Reactions in
Concrete, Mat. Res. Symp. Proc., Vol. 43, 3-20 (1985).
30.36. Plowman, C. and Cabrera, J.G., The Use of Fly Ash to Improve the Sulfate Resistance
of Concrete, Waste Management 16 (1-3), 145-149 (1996).
31.37. Hook, F.E., Morris, E.F., and Rosene, R.B.: Silica-Lime Systems for High Temperature
Cementing Applications, SPE 3447 presented at the 1971 SPE Annual Meeting, New
Orleans, Oct. 3-6.
32.38. Bensted, J.: Microfine Cements, World Cement, December 1992, 45-47 (1992).
33.39. Meek, J.W. and Harris, K.: Repairing Casing Leaks Using Small Particle Size Cement,
III Seminar of Well Cementing, Volume II, 3: Cementing Techniques: Paper 3.2, Caracas,
Intevep S.A. March 18-20, (1992).
34.40. Dahl, J., Harris, K., and McKown, K.: Uses of Small Particles Size Cement in Water
and Hydrocarbon Based Slurries, Proc. 9th Kansas University et al. Tertiary Oil Recovery
Conference, 25-29, (1991).
35.41. Hale, A.H. and Cowan, K.M.: Solidification of Water Based Muds, US Patent 5 058
679 (1991).
36.42. Cowan, K.M., Hale, A.H., and Nahm, J.J.: Conversion of Drilling Fluids to Cements
with Blast Furnace Slag: Performance Properties and Applications for Well Cementing,
paper SPE 24575 presented at the 1992 annual SPE Technical Conference and Exhibition ,
Washington, D.C., October 4-7.
37.43. Lafuma, H.: Expansive Cements, paper presented at the 1952 Intl. Symposium on the
Chemistry of Cement, London.
38.44. Klein, A. and Troxell, G.E.: Studies of Calcium Sulfoaluminate Admixtures for
Expansive Cements, Proc., ASTM (1958) 58, 986-1008.
39.45. Hansen, W.C.: Crystal Growth as a Source of Expansion in Portland-Cement
Concrete, Proc., ASTM (1963) 63, 932-945.
40.46. Expansive Cement ConcretesPresent State of Knowledge, J. American Concrete
Inst. (Aug. 1970) 583.
41.47. Beirute, R. and Tragesser, A.: Expansive and Shrinkage Characteristics of Cements
Under Actual Well Conditions, J. Pet. Tech. (August 1973) 905-09.
42.48. Bied J.: British Patent 8193, (1909).
43.49. Recherches Industrielles sur les Chaux, Ciments et Mortiers, Paris, Dunod, (1926).
44.50. Ciments Lafarge. British Patents 222426 and 222427 (1923).
45.51. Newman, K.: The Design of Concrete Mixes with High Alumina Cement, Reinforced
Concrete Review (March 1960) 5, No. 5.

66
46.52. White, F.L.: Setting Cements in Below Freezing Conditions, Pet. Eng. (Aug. 1952)
B7.
47.53. Maier, L.F. et al.: Cementing Practices in Cold Environments, J. Pet. Tech. (Oct. 1971)
1215-20.
48.54. Morris, E.F.: Evaluation of Cement Systems for Permafrost, paper SPE 2824 presented
at the 1970 AIME Annual Meeting, Denver (February 15-19).
49.55. Bombardieri, C.C., Kijucec, C.C., and Telford, A.S.: Gypsum-Cement Blend Works
Well in Permafrost Area, World Oil (March 1973) 49-52.
50.56. Maier, L.F. et al.: Cementing Materials for Cold environments, JPT, 1215-1220
(1971).
51.57. Bensted, J. Calcium Aluminate Cements: Highlights from a Recent Symposium, World
Cement, October 1990, 452-453 (1990).
52.58. Bensted, J.: Special Oilwell Cements, World Formatted: Indent: Left: 0", Numbered +
Level: 1 + Numbering Style: 1, 2, 3, + Start
Cement, November 1992, 40-45 (1992). at: 7 + Alignment: Left + Aligned at: 0.01" +
53.59. Vereschaka, I.G. et al.: Tsement, No. 10, 10 Tab after: 0.26" + Indent at: 0.26"
(1984).
54.60. Sweatman, R.E. and Scoggins, W.C.: Acid-
Soluble Magnesia Cement: New Applications in Completion and Workover Operations,
paper SPE 18031 presented at the 1980 International Symposium on Oilfield and Geothermal
Chemistry, Stanford, California, U.S.A, May 28-30.
55.61. Tadros, M.E., Skalny, J. and Kalyoncu, R.S.:
Early Hydration of Tricalcium Silicate, J. Am. Cer. Soc., 59, 344-347 (1976).
56.62. Brown, P.W. et al.: Analysis of the Aqueous
Phase During Early C3S Hydration, Cem. Concr. Res., 14, 257-262 (1984).
57.63. Grutzeck, M.W. and Ramachandran, A.R.: An
Integration of Tricalcium Silicate Hydration Models in Light of Recent Data, Cem. Concr.
Res., 17, 164-170 (1987).
58.64. Regourd, M. et al.: Study of Early Hydration of
Ca3SiO5 by X-Ray Photoelectron Spectrometry, Cem. Concr. Res., 10, 223-230 (1980).
59.65. Jennings, H.M.: Aqueous Solubility
Relationships for Two Types of Calcium Silicate Hydrate, J. Am. Cer. Soc., 69, [8], 614-618
(1986).
60.66. Meredith, P., Donald, A.M., and Luke, K.: Pre-induction and Induction Hydration of Formatted: Indent: Left: 0", Numbered +
Level: 1 + Numbering Style: 1, 2, 3, + Start
Tricalcium Silicate:An Environmental Scanning Electron Microscopy Study, J. Mater. Sci., at: 7 + Alignment: Left + Aligned at: 0.01" +
30, 1921-1930 (1995). Tab after: 0.26" + Indent at: 0.26", Tab
61.67. Jennings, H.M and Pratt, P.: An Experimental Argument for the Existence of a stops: Not at 0.26"

Protective Membrane Surrounding Portland Cement During the Induction Period, Cem.
Concr. Res., 9, 501506 (1979).
62.68. Powers, T.C.: Some Physical Aspects of the Formatted: Indent: Left: 0", Numbered +
Level: 1 + Numbering Style: 1, 2, 3, + Start
Hydration of Portland Cement, J. PCA, R &D Labs, 3, 47 (1961). at: 7 + Alignment: Left + Aligned at: 0.01" +
63.69. Birchall, J.D., Howard, A.J., Double D.D.: Tab after: 0.26" + Indent at: 0.26"
Some General Considerations of a Membrane/Osmosis Model for Portland Cement
Hydration, Cem. Concr. Res., 10, 145-155 (1980).
64.70. Stein, H.N. and Stevels, J.M.: Influence of
Silica on the Hydration of 3CaO.SiO2, J. Appl. Chem., 14, 338-346 (1964).
65.71. Brown, P.W., Pommershein, J., and Fronsdorff,
G.: A Kinetic Model for the Hydration of Tricalcium Silicate, Cem. Concr. Res., 15, 35-41
(1985).

67
66.72. Sierra, R.: Investigation of the Hydration
Kinetics of Tricalcium Silicate, 6th ICCC, Vol. 2., Part 1, 138-143 (1976).
67.73. Odler, I. and Drr, H.: Early Hydration of
Tricalcium Silicate II. The Induction Period, Cem. Concr. Res., 9, 277- 284 (1979).
68.74. Fierens, P. and Verhaegen, J.P.: Induction
Period of Hydration of C3S, Cem. Concr. Res., 6, 287-292, (1976).
69.75. Fierens, P. and Verhaegen, J.P.: Hydration of
Tricalcium Silicate in Paste Kinetics of Calcium Ions Dissolution in Aqueous Phase, Cem.
Concr. Res., 6, 337342 (1976).
70.76. Barret, P., Bertrandie, D., and Mntrier, D.:
Comparative Study of C-S-H Formation From Supersaturated Solutions and C3S Solution
Mixtures, 7th ICCC, Vol. 2, 261-265 (1980).
71.77. Young, J.F., Tong, H.S., and Berger, R.L.:
Compositions of Solutions in Contact with Hydrating Tricalcium Silicate Pastes, J. Am.
Cer. Soc. 60, 193-198 (1977).
72.78. Wu, Z.-Q. and Young J.F.: Formation of
Calcium Hydroxide from an Aqueous Suspension of Tricalcium Silicate, J. Am. Cer. Soc.,
67,48-51 (1984).
73.79. Plowman, C. and Cabrera, J.C.: Mechanism and
Kinetics of Hydration of C3A and C4AF Extracted from Cement, Cem. Concr. Res., 14, 238-
242, (1984).
74.80. Pratt, P.L. and Jensen, H.-U.: The Development
of Microstructure During the Setting and Hardening of Cement Pastes, Hydration and
Setting of Cement, RILEM Proceedings 16, Eds. Nonat, A. and Mutin, J.C., 353-360, (1992).
75.81. Tumidajski, P.J. and Thomson, L.: Influence of
Cadmium on the Hydration of C3A, Cem. Concr. Res., 24, 1359-1372, (1994).
76.82. Collepardi, M. et al.: Tricalcium Aluminate
Hydration in the Presence of Lime, Gypsum or Sodium Sulfate, Cem. Concr. Res., 8, 571-
580, (1978).
77.83. Pommersheim, J. and Chang, J.: Kinetics of
Hydration of Tricalcium Aluminate in the Presence of Gypsum, Cem. Concr. Res., 18, 911-
922 (1988).
78.84. Gupta, P., Chatterji, S., and Jeffery, J.W.:
Studies of the Effects of Various Additives on the Hydration Reaction of Tricalcium
Aluminate, Cement Technology, Vol.4, No.2, 63-69, (1973).
79.85. Skalny, J., Jawed I., and Taylor, H.F.W.: Studies
on Hydration of CementRecent Developments, World Cement Technology, Vol. 9,
No.6,183-195 (1978).
80.86. Birchall, J.D., Howard, A.J., and Double, D.D.:
Some General Considerations of a Membrane/Osmosis Model for Portland Cement
Hydration, Cem. Concr. Res. 10, 145-155 (1980).
81.87. Luke, K. and Aitcin, P.-C.: Effect of
Superplasticizer on Ettringite Formation, Advances in Cementitious Materials, Ceramic
Transactions, Vol. 16, 147-166, (1991).
82.88. Luke, K. and Luke, G.: Effect of Sucrose on
Retardation of Portland Cement, Adv. Cem. Res., Vol. 12, No. 1, 9-18 (2000).
83.89. Brown, P.W.: Early Hydration of Tetracalcium
Aluminoferrite in Gypsum and Lime-Gypsum Solutions, J. Am. Ceram. Soc., 70 (7), 493-
496 (1987).

68
84.90. Luke, K. and Glasser, F.P.: Time and
Temperature Dependent Changes in the Internal Constitution of Blended Cements, Il
Cemento, 3, 179-192 (1988).
85.91. Regourd, M.: Crystallization and Reactivity of
Tricalcium Aluminate in Portland Cements, Il Cemento, 75, 323-336 (1978).
86.92. Odler, I. and Wonnemann, R.: Effect of Alkalis
on Portland Cement Hydration. I. Alkali Oxides Incorporated into the Crystalline Lattice
Clinker Minerals, Cem. Concr. Res., 13, 477-482, (1983).
87.93. Older, I. and Wonnemann, R.: Effect of Alkalis
on Portland Cement Hydration. II. Alkalis in the Form of Sulfates, Cem. Concr. Res., 13,
771-777, (1983).
88.94. Jawed, I. and Skalny, J.: Alkalies in Cement: A
Review, II. Effect of Alkalies on Hydration and Performance of Portland Cement, Cem.
Concr. Res., 8, 37-52 (1978).
89.95. Abdul-Maula, S. and Odler, I.: Effect of Oxidic
Composition on Hydration and Strength Development of Laboratory-Made Portland
Cements, World Cement, 13 (5), 216-222 (1982).
90.96. Tang, F.J. and Gartner, E.M.: Influence of
Sulphate Source on Portland Cement Hydration, Adv. Cem. Res., 1, 67-74 (1988).
91.97. Bentur, A.: Effect of Gypsum on the Hydration
and Strength of C3S Pastes, J. Am. Ceram. Soc., 59, 210-213 (1976).
92.98. Jeleni, I., Panovi, A., and Bezjak, A.:
Hydration and Strength Development in Alite-C3A-C S H2-Quartz Pastes Containing
Readily Soluble Alkalies, Cem. Concr. Res., 10, 463-466 (1980).
93.99. Locher, F.W.: Hydration of Pure Portland
Cements, 7th ICCC, Vol.4, 49-61 (1981).
94.100. N Formatted: Indent: Left: 0", Numbered +
Level: 1 + Numbering Style: 1, 2, 3, + Start
egro, A. and Stafferi, Z., Zement-Kalk-Gips, 32, 83 (1979). at: 7 + Alignment: Left + Aligned at: 0.01" +
95.101. T Tab after: 0.26" + Indent at: 0.26", Tab
aylor, H.F.W.: Portland Cement and its Major Constituent Phases, Cement Chemistry, Ed. stops: 0.38", List tab + Not at 0.26"

H.F.W. Taylor, Academic Press Ltd., 31 (1990).

96.102. P
ratt, P.L. and Ghose, A.: Electron Microscope Studies of Portland Cement Microstructures
During Setting and Hardening, Phil. Trans. R. Soc. Lond., A310, 93-103 (1983).
97.103. L
uke, K. and Glasser, F.P.: Selective Dissolution of Hydrated Blast Furnace Slag Cements,
Cem. Concr. Res., 17, 273-282 (1987).
98.104. R
odger, S.A. and Groves, G.W.: Electron Microscopy Study of Ordinary Portland Cement
and Ordinary Portland Cement-Pulverized Fuel Ash Blended Pastes, J. Am. Ceram. Soc., 72
(6), 1037-1039 (1989).
99.105. L
achowski, E.E. et al.: Compositional Development (Solid and Aqueous Phase) in Aged Slag
and Fly Ash Blended Cement Pastes, 10th ICCC, 3ii091, 8pp (1997).
100.106. L
uke, K. and Glasser, F.P.: Chemical Changes Occurring During the Early Hydration of PFA
OPC Mixtures, Mat. Res. Soc. Symp. Proc., Vol. 65 (1986).

69
101.107. L
onguet, P., Burglen, L., and Zelwer, A., La Phase Liquide du Ciment Hydrate, Revue des
Materiaux et Construction, No. 676, 34-41 (1973).
102.108. T
homas, N.L. and Birchall, J.D.: The Retarding Action of Sugars on Cement Hydration,
Cem. Res. Concr., 13, 830-842 (1983).
103.109. D
ouble, D.D., Hellawell, A., and Perry, S.J.: Hydration of Portland Cement, Proc. R. Symp.
London, A 359, 435- 451 (1978).
104.110. O
dler, I., Abdul-Maula, S., and Lu, Z.-Y: Effect of Hydration Temperature on Cement Paste
Structure, Materials Research Society Symposium Proceedings, 85, 139-144 (1987).
105.111. K
jellsen, K.O. and Detwiler, R.J.: Reaction Kinetics of Portland Cement Mortars Hydrated
at Different Temperatures, Cem. Concr. Res., 12, 112-120 (1992)
106.112. B
entur, A. et al.: Structural Properties of Calcium Silicate Pastes: II, Effect of Curing
Temperature, J. Am. Ceram. Soc., 62, 362-366 (1979).
107.113. L
uke, K. and Glasser, F.P.: Effect of Temperature on the Hydration Chemistry and
Durability of Cement Concrete, 4th International Conference on Durability of Materials
and Components, Singapore, 1, 188-195 (1987).
108.114. A
itken, A. and Taylor, H.F.W.: Hydrothermal Reactions in Lime-Quartz Pastes, J. Appl.
Chem., 10, 7-15 (1960).
109.115. B
uckner, D.A., Roy, D.M., and Roy, R.: Studies in the System CaO-Al2O3-SiO2-H2O, II:
The System CaSiO3-H2O, Am. J. Sci., 258, 132-147 (1960).
110.116. A
ssarsson, G.O.: Hydrothermal Reactions of Calcium Hydroxide-Quartz at 120-220, J.
Phys. Chem., 64, 328-331 (1960).
111.117. H
arker, R.I.: Dehydration Series in the System CaSiO3-SiO2-H2O, J. Am. Ceram. Soc., 47,
521-529 (1964).
112.118. F
lint, E.P., McMurdie, H.F., and Wells, L.S.: Formation of Hydrated Calcium Silicates at
Elevated Temperatures and Pressures Research Paper RP 1147, J. Res. of National Bureau
of Standards, 21, 617-638 (1938).
113.119. B
ell, G., Bensted, J., and Glasser, F.P.: Characterization of Hydrothermally-Treated
Calcium Silicate and Oilwell Cement Hydration Products, Adv. Cem. Res., 2, 61-72
(1989).
114.120. P
eppler, R.B.: The System of Lime, Silica and Water at 180C, J. Res. Nat. Bureau of
Standards, 54(4), 205-211 (1955).
115.121. X
i, Y.Z. and Dent Glasser L.S.: Hydrothermal Study in the System Na2O-CaO-SiO2-H2O at
300C, Cem. Concr. Res., 14, 741-748 (1984).

70
116.122. A
ljuvic, B., Bezjak, A., and Glasnovic, A.: Kinetic Study of the Hydrothermal Reaction in
CaO-Quartz System, Cem. Concr. Res., 16, 695-699 (1986).
117.123. L
angton, C.A. et al.: High Temperature Cements with Geothermal Applications, 7th ICCC,
Vol. 3., V/145-V/151 (1980).
118.124. L
uke, K. and Taylor, H.F.W.: Equilibria and Non-Equilibria in the Formation of Xonotlite
and Truscottite, Cem. Concr. Res., 14, 657-662 (1984).
119.125. J
ustnes, H.: Kinetics of Reaction in Cementitious Pastes Containing Microsilica as Studied
by 29Si NMR, Nuclear Magnetic Resonance Spectroscopy of Cement-Based Materials,
Edt., P. Colombet, A.-R. Grimmer, H. Zanni, and P. Sozzani, 245-268 (1998).
120.126. M
ehta, S., Jones, R., Chatterji, J., and McPherson, G.: Effects of Amorphous and Crystalline
Silica on Phase Chemistry, Microstructure and Strength of Set Cement at Elevated
Temperatures, 7th ICCM, 257-270 (1995).
121.127. B
lakeman, E.A. et al.: Studies on the System Sodium Oxide-Calcium Oxide-Silica-Water,
J. Appl. Chem. Biotechnol., 24, 239-245 (1974).
122.128. C
how, S.Y. and Kalousek, G.L.: Research in Cements for Geothermal and Deep Oil Wells,
Paper SPE 5940, Presented at the SPE-AIME 51st Annual Fall Technical conference and
Exhibition, New Orleans Oct. 3-6, 1976 (1976).
123.129. M
itsuda, T. and Taylor, H.F.W.: Influence of Aluminum on the Conversion of Calcium
Silicate Hydrate Gels into 11 Tobermorite at 90C and 120C, Cem. Concr. Res., 5, 203-
210 (1975).
124.130. F
eldman, R.F. and Beaudoin, J.J.: Microstructure and Strength of Hydrated Cement, Cem.
Concr. Res., 6, 389-400 (1976).
125.131. L
uke, K., Taylor, H.F.W., and Kalousek, G.L.: Some Factors Affecting Formation of
Truscottite and Xonotlite at 300-350C, Cem. Concr. Res., 11, 197-203 (1981).
126.132. K
alousek, G.L. and Nelson, E.B.: Hydrothermal Reactions of Dicalcium Silicate and
Silica, Cem. Concr. Res. 8, 283-290 (1980).
127.133. R
oy, D.M. et al.: Potential New High Temperature Cements for Geothermal Wells, paper
SPE 7877 presented at the 1979 International Symposium on Oilfield and Geothermal
Chemistry, Houston, January 22-24.
128.134. N
elson, E.B. and Kalousek, G.L.: Effects of Na2O on Calcium Silicate Hydrates at Elevated
Temperatures, Cem. Concr. Res., 7, 687-694 (1977).
129.135. G
ard, J.A., Mitsuda, T., and Taylor, H.F.W.; Some Observations on Assarsons Z-Phase and
its Structural Relations to Gyrolite, Trucottite and Reyerite, Miner. Mag., 40, 325-333
(1975).

71
130.136. R
oy, D.M. et al.: Hydrothermal High Temperature Cements for Potential Geothermal
Applications, paper SPE 8994 presented at the 1980 International Symposium on Oilfield
and Geothermal Chemistry, Stanford, California, U.S.A, May 28-30.
131.137. L
uke, K., Pulverized Fuel Ash as a Cement Extender, Structure and Performance of
Cements, Edt. J. Bensted and P. Barnes, E&F Spon, 353-357 (2001).
132.138. B
ensted, J.: Oilwell Cements, World Cement, 20, 346-357 (1989).
133.139. M
ehta, P.K.: Sulfate Attack on Concrete: A Critical Review, Concrete Durability, Ed.
R.R.Villarreal, Univ.Autnoma de Nuevo Len, 107-132 (1993).
134.140. T
aylor, H.F.W. : Concrete Chemistry -Carbonation, Cement Chemistry, Ed. H.F.W.
Taylor, Academic Press Ltd., 384-386 (1990).
135.141. E
glinton, M.: Resistance of Concrete to Destructive Agencies Action of Carbon Dioxide,
LeasChemistry of Cemetn and Concrete, 4th edition, Ed. Peter C. Hewlett, Arnold, 329-
334 (1998).
136.142. B
ruckdorfer, R.A.: Carbon Dioxide Corrosion in Oilwell Cements, paper SPE 15176
presented at 1986 Rocky Mountain Regional Meeting, Billings, MT, USA, May 19-21.
137.143. A
ndrade, C. et al.: 8th ICCC, Vol. 5, 256-262 (1986).
138.144. B
ensted, J.: Raman Spectral Studies of Carbonation Phenomena, Cem. Concr. Res. 7, 161-
164 (1977).
139.145. O
nan, D.D. Effects of Supercritical Carbon Dioxide on Well Cements. Paper SPE 15176
presented at the 1986 Rocky Mountain Regional Meeting, Billings, MT, USA, May 19-
21.
140.146. S
ugama, T., Weber, L. and Brothers, L.E.: Sodium-polyphosphate-modified Fly
Ash/Calcium Aluminate Blend Cement: Durability in Wet, Harsh Geothermal
Environments, Materials Letters, 44, 45-53 (2000).
141.147. F
attuhi, N.I. and Hughes, B.P.: The Performance of Cement Paste and Concrete Subjected
to Sulphuric Acid Attack, Cem. Concr. Res., 18, 545-553 (1988).
142.148. S
ugama, T., Weber, T. and Brothers, L.E.: Resistance of Sodium Polyphosphate-modified
Fly Ash/Calcium Aluminate Blend Cements to Hot H2SO4 Solution, Cem. Concr. Res.,
29, 1969-1976 (1999).
143.149. L
ota, J.S. et al.: Hydration of Class G Oilwell Cement at 20C and 5C, LIndustria
Italian, del Cemento, No.725, 776-798 (1997).
144.150. T
noutasse, N.: The Hydration Mechanism of C3A and C3S in the Presence of Calcium
Chloride and Calcium Sulfate, Proc., Fifth Intl. Cong. Chem. Cement, Paris Pt. 2, 372
378 (1978).

72
145.151. S
hell, F.J. et al.: Low Density Oil Well Cements, World Oil, 143(4) 131-140, (1956).
146.152. S
mith, R.C. and Calvert, D.G.: The Use of Sea Water in Well Cementing, J. Pet. Tech.,
759-764 (1975).
147.153. P
ietersen, H.S., Fraay, A.L.A., and Bijen, J.M.: Reactivity of Fly Ash at High pH, Mat.
Res. Soc. Symp. Proc. Vol. 178, 139-157 (1990).
148.154. Y
oung, J. F.: Reaction Mechanism of Organic Admixtures with Hydrating Cement
Compounds. Transportation Res. Record, 564: 1 9 (1976).
149.155. K
halil, S.M. and Ward, M.A.: Influence of a Lignin Based Admixture of the Hydration of
Portland Cements, Cem. Concr. Res., 3, 677-688 (1973).
150.156. R
amachandran, V.S.: Interaction of Calcium Lignosulfonate with Tricalcium Silicate,
Hydrated Tricalcium Silicate and Calcium Hydroxide, Cem. Concr. Res., 2, 179-194
(1972).
151.157. C
ollepardi, M., Marcialis, A., and Solinas, V.: The Influence of Calcium Lignosulfonate on
the Hydration of Cements, Il Cemento, 70, 3-14 (1970).
152.158. J
ennings, H.M. et al.: Interpretation of the Effects of Retarding Admixtures on Pastes of
C3S, C3A Plus Gypsum and Portland Cement, Proc. 8th ICCC, 239-243 (1986).
153.159. Y
oung, J.F.: Hydration of Tricalcium Aluminate with Lignosulfonate Additives, Mag.
Concr. Res., 14, 137-142 (1962).
154.160. J
awed, I., Klemm, W.A., and Skalny, J.: Hydration of Cement-Lignosulfonate-Alkali
Carbonate System, J. Amer. Ceram. Soc., 62, 461-464 (1979).
155.161. D
urekovic, A.: Hydration in Very High Strength Cement Composites, 9th Inter. Congr.
Chem Cement, New Delhi, pp. 543 549 (1992).
156.162. T
enoutasse, N. and Sing, N.B.: Effect of Glucose and Calcium Gluconate on the Hydration
of Portland Cement, Indian, J. Technol, 16(5), 184-189 (1978).
157.163. M
onosi, S. et al., Influence of Lignosulfonate, Glucose and Gluconate on the C3A
Hydration, Cem. Concr. Res., 13, 568-574 (1983).
158.164. D
iamond, S.: Interaction Between Cement Minerals and Hydroxycarboxylic Acid
Retarders, J. Amer. Ceram. Soc., 55, 405-408 (1972).
159.165. B
eaudoin, J.J. and Ramachandran, V.S.: Physico-Chemical Characteristics of Low Porosity
Cement Systems, Ch. 6, Material Science of Concrete, Vol. III, pp. 362, The American
Ceramic Soc. (1992).
160.166. D
ouble, D.D.: New Developments in Understanding the Chemistry of Cement Hydration,
Trans. Royal Soc. London Ser. A 310, 53 66 (1983).

73
161.167. M
ilestone, N.B.: Hydration of C3S in the Presence of Lignosulfonates, Glucose and Na-
Gluconate, J. Amer. Ceram. Soc., 62, 321-324 (1979).
162.168. R
amachandran, V.S. et al.: The Role of Phosphonates in the Hydration of Portland
Cement, Materials and Structures, 26, 425-432 (1993).
163.169. L
iebour, W.: The Influence of Lead and Zinc Compounds on the Hydration of Portland
Cements, Proc. 5th Int. Symp. Chemistry of Cements, 2, 444-453 (1968).
164.170. A
rliguie, G., Ollivier, J.P., and Grandet, J.: Etude de lEffect Retardateur du Zinc sur
LHydratation de la Pate de Ciment Portland, Cem. Concr. Res., 12, 79-86 (1982).
165.171. R
amachandran, V.S., Seeley, R.C., and Polomark, G.M.: Free and Combined Chloride in
Hydrating Cement and Cement components, Materials and Structures, 17, 285-289
(1984).
166.172. S
tandard Specification C618, Annual Book of ASTM Standards, Vol. 4.02, American
Society for Testing and Materials, West Conshohocken, Pennsylvania (2000).
167.173. M
cCarthy, G.J. et al.: Use of a Database of Chemical, Mineralogical and Physical Properties
on North American Fly Ash to Study the Nature of Fly Ash and Its Utilization as a Mineral
Admixture in Concrete, Mat. Res. Soc. Symp. Proc., 178, 3-34 (1990).
168.174. D
iamond, S.: On the Glass Present in Low-Calcium and in High-Calcium Flyashes, Cem.
Concr. Res., 13, 459-464 (1983).
169.175. S
cheetz, B.E. et al.: Physical and Chemical Behavior of Selectively Etched Fly Ashes,
Mat. Res. Soc. Symp. N, 24-33 (1982).
170.176. G
roves, G.W. and Richardson, I.G.: Microcrystalline Calcium Hydroxide in Pozzolanic
Cement Pastes, Cem. Concr. Res., 24, 1191-1196 (1994).
171.177. D
iamond, D.: Effects of Microsilica (Silica Fume) on Pore Solution Chemistry of
Cements, Communication of Amer. Ceram. Soc., C82-84 (1983).
172.178. T
enoutasse, N. and Marion, A.M.: The Influence of Silica fume in Alkali-Aggregate
Reactions, Concrete Alkali-Aggregate Reactions, Ed. P.E. Grattan-Bellew, 71-75 (1987).
173.179. H
uang, C.-Y. and Feldman, R.F.: Influence of Silica Fume on the Microstructural
Development in Cement Mortars, Cem. Concr. Res., 15, 285-294 (1985).
174.180. O
dler, I. and Becker, Th.: Effect of Some Liquefying Agents on the Properties and
Hydration of Portland Cement and Tricalcium Silicate Pastes, Cem. Concr. Res., 10, 321-
331 (1980).
175.181. A
saga, K. and Roy, D.M.: Rheological Properties of Cement Mixes: IV. Effect of
Superplasticizers on Viscosity and Yield Stress, Cem. Concr. Res., 10, 287-295, (1980).

74
176.182. M
ichaux, M., Oberste-Padtberg, R., and Defosse, C.: Oilwell Cement SlurriesPt. 2:
Adsorption Behaviour of Dispersants, Cem. Concr. Res., 16, 921-930 (1986).
177.183. A
nderson, P.J., Roy, D.M., and Gaidis, J.M.: The Effects of Adsorption of Superplasticizers
on the Surface of Cement, Cem. Res. Concr., 17, 805-813 (1987).
178.184. H
anna, E. et al.: Rheological Behaviour of Portland Cement in the Presence of a
Superplasticizer, Third CANMET/ACI International Conference on Superplasticizers and
Other Chemical Admixtures in Concrete, SP 119-9, 171-188 (1989).
179.185. S
haughnessy, R. III and Clark, P.E.: The Rheological Behavior of Fresh Cement Pastes,
Cem. Concr. Res., 18, 327-341 (1988).
180.186. C
ollepardi, M., et al.: Influence of Sulphonated Naphthalene on the Fluidity of Cement
Pastes, 7th ICCC, Vol. III, VI-20 VI-25 (1980).
181.187. S
tephens, M. and Patel, B.B.: Well Cement Slurries and Dispersants Therefor, U.S. Pat.
4,923,516 (1990).
182.188. E
off, L.: Acetone/Formaldehyde/Cyanide Resins, U.S. Pat. 5,290,357 (1994).
183.189. R
ixom, M.R.: Chemical Admixtures for Concrete, (E. & F.N. Spon Ltd., London), 1-89
(1978).
184.190. L
uke, K.: Influence of Superplasticizer on Aqueous Phase Composition in the CaO-Al2O3-
SO3-H2O System, Adv. Cem. Res., Vol.4, No.15, 103-110 (1992).
185.191. R
amachandran, V.S.: Adsorption and Hydration Behavior of Tricalcium Aluminate-Water
and Tricalcium Aluminate-Gypsum-Water Systems in the Presence of Superplasticizers,
3rd International Congress on Polymers in Concrete, 1071-1081 (1981).
186.192. U
chida, S. et al.: Influence of Characters of Cement on Fluidity of Fresh Paste and Mortar
with Organic Admixtures, Journal of Research, Onoda Cement Company, 42, 123, 4-15
(1990).
187.193. B
urk, A.A., Gaidis, J.M., and Rosenberg, A.M.: Adsorption of Naphthalene-Based
Superplasticizers on Different Cements, II Inter. Conf. Superplasticizers in Concrete, 23
(1981).
188.194. C
unningham, J.C., Drury, B.L. and Gregory, T.: Adsorption Characteristics of Sulphonated
Melamine Formaldehyde Condensates by High Performance Size Exclusion
Chromatography, Cem. Concr. Res., 19, 919-926 (1989).
189.195. A
ndersen, P.J., Roy, D.M., and Gaidis, J.M.: The Effect of Superplasticizer Molecular
Weight on Its Adsorption on, and Dispersion of, Cement, Cem. Concr. Res., 18, 980-
986(1988).
190.196. E
ncyclopedia of Polymer Science and Engineering, Volume 11, John Wiley & Sons (1988)

75
191.197. S
aunders, K.J., Organic Polymer Chemistry, Chapman and Hall (1973).
192.198. G
olding, B: Polymers and Resins, Their Chemistry and Chemical Engineering, D. Van
Nostrand Company, Inc. (1994).
193.199. B
ray, W.S. and Wood, W.R.: Well Cementing Method Using a Dispersant Intensifier, U.S.
Pat. No. 5,105,885 (1992).
194.200. C
arpenter, R.B.: Matrix Control Cementing Slurry, U.S. Pat. No. 4,569,395 (1986).
195.201. G
rulke, E.A.: Polymer Process Engineering, PTR Prentice Hall.
196.202. R
amachandran, V.S.: Concrete Admixtures Handbook Properties, Science, and Technology,
2nd Edition, Noyes Publications (1995).
197.203. P
arcevaux, P.A. et al.: Cement Compositions for Cementing Wells, Allowing Pressure Gas-
Channeling in the Cemented Annulus to be Controlled, U.S. Pat. No. 4,537,918 (1985).
198.204. C
hilds, J.D. and Burkhalter, J.F.: Fluid Loss Reduced Cement Compositions, UK Pat. No.
2,247,234 (1992).
199.205. H
ook, F.E.: Aqueous Cement Slurry and Method of Use, U.S. Pat. No. 3,483,007 (1969).
200.206. S
hell, F.J. and Wynner, R.A.: Applications of Low-Water-Loss Cement Slurries, Paper
API 875-12-1 (1958).
201.207. G
reminger, G.K.: Hydraulic Cement compositions for Wells, U.S. Pat. No. 2,844,480
(1958).
202.208. W
eisend, C.F.: Composition Comprising Hydroxyethyl Cellulose, Polyvinyl Pyrrolidone,
and Organic Sulfonate, Cement Slurry Prepared Therefrom and Method of Cementing Well
Therewith, U.S. Pat. No. 3,132,693 (1964).
203.209. C
hatterji, J. et al., Liquid Water Loss Reducing Additives for Cement Slurries, U.S. Pat.
No. 4,433,731 (1984).
204.210. M
organ, R.L. and Kolodny, E.R.: Hydraulic Cement Compositions Containing Acrylamide-
Acrylic Acid Copolymers and Method of Making Same, U.S. Pat. No. 2,868,753 (1959).
205.211. M
cKenzie, L.F. and McElfresh, P.M.: Acrylamide/Acrylic Acid Copolymers for Cement
Fluid Loss Control, paper SPE 10623 presented at the 1982 International Symposium on
Oilfield and Geothermal Chemistry, Dallas, January 25-27.
206.212. A
dams, S.L. et al.: Low Fluid Loss Cementing Compositions Containing
Polyvinylpyrrolidone-Polyacrylamide Triblock Polymers and Their Use, U.S. Pat. No.
3,994,852 (1976).
207.213. P
ersinski, L.J. et al.: Low Fluid Loss Cementing Compositions Containing Hydrolyzed

76
 Acrylamide/2-Acrylamido-2-Methylpropane Sulfonic Acid Derivative Copolymers and
Their Use, U.S. Pat. No. 4,015,991 (1977).
208.214. N
ewlove, J.C. et al.: Fluid Loss Control in Oil Field Cements, U.S. Pat. No. 4,480,693
(1984).
209.215. S
edillo, L.P. et al.: Fluid Loss in Oil Field Cements, U.S. Pat. No. 4,659,750 (1987).
210.216. M
cKenzie, L.F. et al.: Non-Retarding Fluid Loss Additives for Well Cementing
Compositions, U.S. Pat. No. 4,602,685 (1986).
211.217. R
ao, S. P. et al.: Hydrolytically Stable Polymers for Use in Oil Field Cementing Methods
and Compositions, U.S. Pat. No. 4,555,269 (1985).
212.218. B
rothers, L.E.: Method of Reducing Fluid-Loss in Cement Compositions Containing
Substantial Salt Concentrations, U.S. Pat. No. 4,640,942 (1987).
213.219. F
ry, S.E. et al.: Method of Reducing Fluid Loss in Cement Compositions Which May
Contain Substantial Salt Concentrations, U.S. Pat. No. 4,676,317 (1987).
214.220. H
ille, M. et al.: Cement Slurries for Deep Holes, with a Copolymer Content for Reducing
the Water Loss, U.S. Pat. No. 4,587,283 (1986).
215.221. D
efosse, C.: Fluid Loss Additive for Cement, U.S. Pat. No. 4,568,471 (1986).
216.222. B
rothers, L.: Method of Reducing Fluid Loss in Cement Compositions Which May Contain
Substantial Salt Concentrations, U.S. Pat. No. 4,700,780 (1987).
217.223. S
avoly, A. et al.: Fluid Loss Agents for Oil Well Cementing Compositions, U.S. Pat. No.
4,674,574 (1987).
218.224. S
tephens, M.: Fluid Loss Additive for Cement Slurries Containing a N-Vinyl-2-Pyrrolidone
/ 2-Acrylamido-2-Methylpropane Sulfonate / Acrylic Acid / Acrylamide Polymer, U.S.
Pat. No. 5,109,042 (1992).
219.225. S
tephens, M.: Fluid Loss Additives for Well Cementing Compositions Containing a
Tetrapolymer, U.S. Pat. No. 5,153,240 (1992).
220.226. P
eiffer, D.G. et al.: Fluid Loss Control in Oil Field Cements, U.S. Pat. No. 4,626,285
(1986).
221.227. H
uddleston, D.A. et al.: Method for Reducing Fluid Loss From Oilfield Cement Slurries
Using Vinyl- Grafted Wattle Tannin, U.S. Pat. No. 5,134,215 (1992).
222.228. W
eisend, C.F.: Cement Additive Containing Polyvinylpyrrolidone and a Condensate of
Sodium Naphthalenesulfonate with Formaldehyde, U.S. Pat. No. 3,359,225 (1967).
223.229. C
owan, K.M. and Eoff, L.: Surfactants: Additives To Improve the Performance Properties

77
 of Cements, paper SPE 25181 presented at the 1993 International Symposium on Oilfield
Chemistry, New Orleans, March 2-5.
224.230. S
cott, P.O. Jr., Lummus, J.L., and Howard, G.C.: Methods for Sealing Vugular and
Cavernous Formations, Drill. Contractor, 70-74 (1953).
225.231. G
oins, W.C. Jr.: Lost Circulation Problems Whipped with BDO (Bentonite Diesel Oil)
Squeeze, Drilling, 15 No.11, 83 (1954).
226.232. M
essenger, J.U. and McNeil, J.S. Jr.: Lost Circulation Corrective: Time Setting Clay
Cement, Trans., AIME, 195, 171-182 (1952).
227.233. H
oward, Q.C. and Scott, P.P. Jr.: An Analysis and Control of Lost Circulation, Trans.,
AIME, 192, 171-182 (1951).
228.234. W
hite, R.J.: Lost Circulation Materials and Their Evaluation, Drill. And Prod. Practice,
API, 352-359 (1956).
229.235. C
hilds, J., Sabins, F., and Taylor, M.J.: Method of Using Thixotropic Cements for
Combating Lost Circulation Problems, U.S. Patent 4,515,216 (1985).
230.236. E
ilers, L.H., Nelson, E.B., and Moran, L.K.: Journal of Petroleum Technology, 35, 1373
(1983).
231.237. B
our, D.L., Daugherty, D., and Sutton, D.L.: New Expansive Cement System for High
Temperature, Proc. Southwestern Petroleum Short Course, Lubbock, TX (1988).
232.238. S
pangle, L.B.: Expandable Cement Composition, European Patent No. 254,342 (1988).
233.239. C
arter, L.G., Waggoner, H.F., and George, C.R.: Expanding Cements for Primary
Cementing, JPT, 551-558 (1966).
234.240. B
each, H.J. and Goins, W.C. Jr.: A Method of Protecting Cements Against Harmful Effects
of Mud Contamination, Trans. AIME, 210, 148-152 (1957).
235.241. C
arter, L.G., Slagle, K.A., and Smith, D.K.: Resilient Cement Decreases Perforating
Damage, paper presented at API Mid-Continent Dist. Div. of Production Spring Meeting,
Amarillo (1968).
236.242. M
ueller, D.T., Virgillio, G., and Dickerson, J.P.: Stress Resistant Cement Compositions and
Methods Using Same, U.S. Patent 6,230,804 (2001).
237.243. M
ueller, D.: An Evaluation of Well Cements for Use in High Strength Environments,
Harts Petroleum Engineer International, April 1988 (1988).
238.244. A
STM Standard Methods of Sampling and Testing Calcium Chloride for Roads and
Structural Application, Annual Book ASTM Standards, Pt 14, 601-602 (1980).
239.245. U
chikawa, H. and Furuta, R.: Application of High Performance Liquid Chromatography to

78
 the Analysis of Organic Additives in Hardened Cement Paste, Onodo Kenkyu Hokoku,
31, (101) 1-6 (1979): Chem Abstr., 92, 202541 (1980).
240.246. M
uszynski, L.C.: Analysis of Hardened Concrete for Admixture Content, Purdue
University, Report PB80-122294, 142 pp, (1979).
241.247. C
onnolly, J.D., Hime, W.G., and Erlin, B.: Analysis of Admixtures in Hardened Concrete,
Proc. Int. Congr. Admixtures, London 1980, Construction Press, 114-129 (1981).
242.248. N
ishi, S. and Hayashi, M.: Determination of Total Lignsosulfonate in Cement, Hardened
Mortar and Hardened Concrete, Rev. 33rd Gen. Mtg. Cement Assoc. Japan (Engl. Abstr.)
89-90 (1979).
243.249. F
unato, M., Nakashima Y., and Nakajima, K.: Determination of -Naphthalene Sulfonate
Formaldehyde Condensate (BNSF) Water-Reducing Admixture in Hardened Cement
Mortars by High Performance Liquid Chromatography, Sem. Gijutsu Nempo, 35, 94-97
(1981).
244.250. G
ross, D. and Rooss, H.: HPLC Study of Admixtures for Use in Concrete, Betorwerk
Fertigfiel.-Tech., 49(5) 17-312 (1983) Chem Abs 99, 92815 (1984).
245.251. R
eul, H.: Determination of Melment in Cement Mixtures by UV Spectrophotometry, Z.
Anal. Chem., 285, (2) 605-610 (1977).
246.252. O
no, A., Tai, Y., and Inowaka, H.: Qualitative and Quantitative Analysis of Organic
Admixtures Contained in Concrete, Semento, Knokurito Ronbunshu, 47,232-237 (1993).
247.253. M
aruta, T. and Kato, K.: Determination of Organic Admixtures in Hardened Cement Paste,
Sem. Gijutsu Nempo, 32, 101-104 (1977): Chem Abstr. 92, 10293 (1980).

79