You are on page 1of 20

Ship Configurations and Insulation

Design / Application
by
Gordon H. Hart, P.E., Artek, Inc.
Pat Fulton, Silencing Ocean Systems
Gerald Cox, Performance Contracting, Inc.

Abstract
Since their early development, the construction and outfitting of steel vessels have
presented a unique challenge to the insulation designer in ensuring comfort and quality
insulation treatments. The drive to make large commercial and military sea-going vessels
lighter, faster, and stronger invariably contributes to complexities of stiffening members,
compartmentalization, and system integration. In so doing, the designer must first balance
the cost of thermal insulation treatments against several competing factors: the capacity
of heating and cooling equipment, the cost of this equipment, and the cost of energy per
unit time to meet thermal requirements.
In the past, the US shipbuilding industry has relied on a fixed table of maximum
allowable thermal transmittance values, or U values, to determine the thickness of
insulation for particular configurations. In this paper, the authors show that these U
values are inadequate, in comparison to current standards for the use of thermal insulation
on walls and envelopes in building construction, and that a selective increase in insulation
thicknesses used on ships can reduce the weight of fuel and equipment for space heating
and cooling. The authors also propose that the insulation designer be encouraged to
incorporate different methods of estimating heat flows given specific environmental
conditions and stiffener configurations compared with long-held industry standards.
These methods include computer-assisted Finite Element Analysis, recognition of
varying extreme conditions, and actual stiffener configurations that contribute to thermal
flows. With these changes, the insulation systems for US built ships could be improved
thermally, the total ship weight could be reduced, and the insulation systems could be
installed more quickly, thereby reducing the cost of construction.

Introduction
The steel boundaries of ocean going ships are thermally insulated for basically the same
reasons we insulate the envelopes of any habitable structure: to reduce energy use for
space heating and cooling, to provide an acceptable thermal environment for occupants,
to prevent moisture condensation on the interior surfaces, to protect against the spread of
fire, and / or to protect stored goods and materials and, in some instances, insulation is
required to mitigate noise migration for occupants. The piping and mechanical
equipment on ships are insulated for many of the same reasons: to reduce process or
HVAC related energy use, to reduce loads on mechanical equipment, to provide
personnel protection, and / or to reduce noise. Increasingly, we insulate systems to
reduce quantities of emissions and thereby reduce the adverse impact on the environment.

For over 40 years the American ship construction industry has relied on a document first
published by the Society of Naval Architects and Marine Engineers (SNAME) in 1963
for the selection of structural insulation thicknesses. While thermal insulation standards
in residential, commercial buildings, and industrial environments have changed
dramatically in those 40 years due to both significantly higher energy prices and to a
greater understanding of insulation performance, those standards for the American
marine industry have not changed.

This paper will review the thermal requirements for marine insulation on ship boundaries,
the methodology for determining insulation thicknesses, some of the problems associated
with using previously prescribed thicknesses, and then make recommendations for
changes.

Reasons for Insulating Ships Shells,


Bulkheads, and Decks
From the perspective of requirements for thermal insulation, a ship can be compared to
an entire community comprised of different types of structures. There is insulation for
the thermal envelopes, which separates indoors from outdoors, such as one would see in
houses or commercial buildings. There is also mechanical insulation on piping and air

handling ducts, such as one might see in a typical commercial building, as well as
insulation for fire proofing and fire barriers and for noise reduction. Finally, there is
mechanical insulation such as one might see at an industrial facility such as an electric
power generating plant.

Crewmembers occupy or work in a variety of areas aboard ship. Whether living spaces,
cargo spaces, or workspaces, the various compartments within a ship need to meet certain
acceptable criteria for either human comfort or for acceptable working conditions. To
maintain inside air temperatures within an acceptable range, heating and cooling
systems must be designed in combination with thermal insulation systems. In stowage
spaces, maintaining air temperatures and moisture levels within predetermined limits
often protect cargos. In machinery and equipment spaces, temperatures and moisture
levels must likewise be kept within acceptable limits. Frequently, within a ship, one
cargo space or mechanical equipment space may need to be kept at a different
temperature than an adjacent space, requiring structural insulation to thermally separate
the compartments.

The shell boundaries and decks of ships are critical paths of heat flow. They separate the
interior, temperature and humidity conditioned space from the outside weather and sea
conditions. Parts of the hull above water are exposed to ambient air, and the parts below
water, exposed to the ocean water. For the portions of the hull that are below water level,
because the water temperature can be much lower than the inside air temperature and
because the heat transfer coefficient from water is much greater than for air, insulation
must both limit heat loss to the water and prevent moisture condensation on the cold steel
surfaces. Heat transfer through a ships shell is most extreme where shell plate stiffeners
act as heat transfer fins which, if not adequately insulated, significantly increase the
overall heat transmission. The heat flow through these interfaces is one of the focal
points of this paper. The typical minimum design outdoor air temperature for heating
conditions is 0 F; the typical maximum design exterior surface for cooling conditions is
135 - 155 F, values which account for a combination of both a hot air temperature and
strong incident solar radiation.

In a stationary environment, such as a shore side structure, the ambient conditions


typically change over the course of a year, with the expected ambient conditions
depending on both location and time of year. In the case of a ship, traveling the globe to
different climatic conditions those ambient conditions are in a constant state of change
affected by both location of the ship and time of year. A modern ships envelope and
heating and cooling systems must be designed for the extremes of those ambient
conditions.

Shipboard generator units, usually diesel driven, are typically used to generate electricity
on a ship. This electricity is used to power both resistance space heaters as well as vaporcompression air-cooling systems. A ships electrical generating load capacity is a
primary concern in the design and operation of any vessel. That capacity is always one
of the critical factors of a ships performance specifications. These diesel generators, of
course, require diesel fuel to operate, and this fuel must be carried within the ship, adding
to the total weight of the ship. Furthermore, the diesel generators must be sized
adequately to provide sufficient electrical power for both cold conditions, in polar or artic
zones of the earth, and hot conditions in the tropics. The greater the heating and cooling
load, the larger the generators and the greater the amount of diesel fuel, the greater the
added weight to the overall weight of the ship. In the ship construction industry, weight
limitation is a very important design consideration. Therefore, there should be a strong,
inverse relationship between the amount and location of installed thermal insulation and
the total combined weight of the diesel generators and the diesel fuel to power them.
This inverse relationship is an extremely important issue that cannot be overemphasized.

Scantling Effects
Scantling and plate stiffener configuration greatly affect heat flows and heating and
cooling loads. Modern ships are structurally reinforced with a variety of stiffeners. To
provide an effective thermal envelope, one with a low U value, the thermal insulation
must be placed against the steel boundaries and around the steel stiffeners, facing the
conditioned space. Figures 1 and 2 show uninsulated and insulated stiffeners. When
conditions warrant, the industry attempts to design the insulation for the unstiffened side
of an internal bulkhead since that surface is planar and hence easier to insulate.

To understand the heat transmittance through a complex steel boundary, we must


consider the relationship between a number of variables. These include the thickness of
the plate, the distance between stiffeners, the depth, thickness, and configuration of the
stiffeners, the insulating value of the insulation and its applied thickness on both the
plane and stiffened surfaces. Each of these factors must be examined and accounted for
in the determination of a composite U value for each assembly or boundary.

Figure 1: A close-up of a partially insulated stiffener on the hull of a ship.

Figure 2: A view of several stiffeners on a ship, some having been insulated and some
waiting to be insulated.
.

Historical Insulation Treatment


The standard thermal insulation material used for exposed structural boundaries such as
the shell, bulkheads, and decks is referred to as hull board or Navy Board. The
standard military specification for thermal insulation material for US Navy vessels hulls
is MIL-I-742. This is a semi-flexible, nominal 2.8-lbs/ft3 fibrous glass wool board with
3.0% organic binder content. These boards are typically faced with a heavy-duty, very
durable, coated, woven fiberglass cloth, or reinforced Mylar facing. Standard practice
requires that this product be installed on the ship hulls using weld pins, washers, and / or
drive-on caps. A typical shell boundary is shown in the extract from a contemporary
midship section, Figure 3A. Shell stiffening on modern ships often utilize extensive bulb
plate stringers. For structural reasons, based on ship vibration, weld pins are typically
installed with 12 to 18 inch spacing and no more than 6 inches from the edge of a board.
A matching woven glass fiber tape is applied over the facing joint of two adjoining
boards and at all corners and thermal edges.

As illustrated here, common sizes for shell longitudinal stiffeners run from 260 x 12 mm
to 300 x 13 mm bulb plate.

While there are options available in the selection of thermal insulation on a typical ship,
this paper is going to focus on the insulation criteria and methodology for the selection
and installation of semi-rigid fibrous glass insulation boards on structural boundaries
such as the ships hull or shell, bulkheads, and decks.

SNAME Technical and Research Bulletin 4-7 (1963)

The standard for determining hull insulation thicknesses, for both US built commercial
vessels and Navy ships built to commercial standards, is given in Reference 1: SNAME
Technical & Research Bulletin 4-7, Thermal Insulation Report (hereafter referred to
simply as SNAME 4-7). The data contained in this publication was generated by a series
of thermal tests that were conducted to determine insulation performance in combination
with two-dimensional angle stiffeners. This paper challenges that basic premise and the
applicability of SNAME 4-7 to todays ship construction. Many boundaries on todays

ships are stiffened with bulb type stiffeners, which create differing heat flow for similar
insulation thickness. Additionally, the authors believe that the maximum allowable
thermal transmittance values set forth in the SNAME 4-7 are too high to permit efficient
space heating and cooling given current and future costs of energy.
In SNAME 4-7, the thermal requirements for hulls, bulkheads, and decks are given in
terms of maximum allowable thermal transmittance values, or U values. Insulation
thicknesses are then derived from those values. These maximum allowable U values
are given in the SNAME document in a Table 2 titled Temperature differences versus
U factors. This table is shown below:

Table 2 from the SNAME Technical & Research Bulletin 4-7:


Temperature Differences vs. U Values
Temperature Differences ( F)

Maximum U values
(Btu/hr - ft2 - F)

0 to 15
16 to 30
31 to 50
Over 50

1.75
0.37
0.26
0.16

To use this table to determine hull insulation thickness, one must first be given the design
Temperature Difference for the particular section of insulation. Using other tables in the
research report, he then determines the Surface Coefficient values for the particular air
temperature difference, direction of airflow, and location of the air film. He then
determines the thickness using still another set of tables, based on those thermal tests
conducted in the early 1960s, which account for additional heat loss due to the presence
of the rib stiffeners. These insulation thickness tables go on for many pages, with
different sets for insulation thickness and choices for direction of heat flow, winter or
summer conditions, and boundary conditions (inside air to outside air, inside air to sea
water, inside air to inside air). The tables are based on 36 stiffener spacings for angle
stiffeners measuring 6 x 4. Since there are numerous other stiffener spacings, angle
stiffener sizes, and stiffener designs used in ship construction, this gives the reader an
insight into the difficulties of using these tables for insulation thickness determination.
There are, in fact, other tables for additional stiffener spacings than 36 in the SNAME
document and these can be used to adjust the U values obtained from the insulation
thickness tables. However, because they are premised on a single size of a single angle
type of stiffener, they cannot account accurately for the U values for the wide variety of

stiffener design variables found in the decks, bulkhead assemblies, and other thermal
boundaries encountered even within one ship.

A Practical Example
For example, to illustrate a design problem, let us say that we are considering a portion of
the hull that separates heated inside air at 60 from cold outside air at 0 during the
Winter Conditions, giving a design Temperature Difference of 60 F. In this instance,
there is no lining separating the hull from the indoor air. So, the Table 2 above shows our
maximum U value allowed, for a Temperature Difference for over 50, of 0.16 (at this
point, the units are understood and will not be repeated). From experience, we know that
the ends of the angle stiffeners will have to be covered with insulation; it will not be
enough to simply insulate the plane surfaces of the hull. We flip through the pages of
tables looking for the configuration just described. On page 76, Table 17, Type 52, we
find one that meets the maximum U value allowable of 0.16 (see Attachment 1 for a
reproduction of this page). For Horizontal Heat flow, Winter conditions, for Inside Air
to Weather Air, with 2 of insulation board on the hull and 1 completely over the angle
stiffeners, the table gives us a U value of 0.131. Now, the shipbuilder may consider
that he could save some material and labor by not insulating the angle stiffeners to that
degree. However, if he were to leave the stiffeners uninsulated, this would obviously
result in a higher U value for that assembly. To determine the attendant U value, one
would look on the same page at Type 50, which has the same 2 on insulation board on
the hull but none on the angle stiffener. There, he can see that his U value would be
0.326, a value that greatly exceeds 0.16. Therefore, for this set of design conditions, the
contractor will have to insulate the angle stiffeners as well as the flat surfaces between
the stiffeners to reduce the U value below 0.16.

As an example of heat that flows through stiffeners, the following extraction from
SNAME 4-7 illustrates the heat mitigation value of insulation covering the stiffener:

Note that the heat transmittance (U-value) decreases more than 57 % with the
introduction of insulation covering the entire stiffener. It can also be recognized that heat
flows primarily through stiffener members of the ships structure.

For those who are accustomed to using computer programs such as 3E-Plus, this SNAME
4-7 methodology is reminiscent of how things used to be done in the design-engineering
world. We would remind the reader, however, that this SNAME 4-7 document, while
dated, is incredibly comprehensive and detailed for the subject matter it is addressing,
namely two-dimensional heat transfer through a combination of steel and fibrous glass
board insulation, with two fluid boundary conditions. The most practical way to improve
on this methodology with modern technology is with Finite Element Analysis (FEA), a
heat transfer methodology for which there are very few practitioners in the insulation
industry. Further, while it is commendable that this SNAME 4-7 methodology takes
account for two-dimensional system thermal performance, it does not allow for variations
in system design that might provide an equal thermal performance with either less
insulation or with a less labor-intensive application. Therefore, FEA may be the only
way we can improve on this SNAME 4-7 methodology (note that 3E-Plus will not do the
job accurately since it assumes one-dimensional heat transfer and therefore cannot
account for heat transfer through the two-dimensional angle stiffeners).

Returning to the SNAME 4-7 document Table 2 above, one needs to ask whether these
maximum allowable U values are adequate for todays energy costs and practices. A
U value of 0.16 is equivalent to a system R value of 1 / 0.16 = 6.25 F-hr-ft2 / Btu.
Many of us are accustomed to seeing R value recommendations, such as from the
Department of Energy, for insulating walls and ceilings of houses. These are not always
system R values; rather, they are typically limited to those portions without doors and
windows and without considering the effects of studs. So, for a comparison, let us
consider a cold climate location such as central Indiana , which has 5700 Heating Degree
Days. There, DOE insulation calculator2 recommends, for a wood-framed house with
natural gas heating, that attic ceiling insulation have an R value of 49 and the above
ground walls have a combination of R-5 sheathing and R-13 between the studs, for a total
R value through the insulation of R-18. Taking a cursory look at this one example
using the DOE standard calculator, one can see that the insulation requirements for the
Navy ships, per the SNAME 4-7 document, are significantly less. We might need
insulation thicknesses on the order of 6 inches or so to achieve the same R values
typically installed in house walls in cold climates. So, it would appear that the SNAME
4-7 thickness requirements, essentially given in Table 2, have not kept pace with todays
insulation practices.

Another consideration is with the assumptions behind the SNAME 4-7 design
methodology. The issues that skew the thermal performance of stiffened, insulated
boundaries are the spacing and the configuration of the stiffeners. The data in SNAME
4-7 is based on using angle stiffeners measuring 6 x 4. Tests, conducted more than
four decades ago and using 4 x 3 x thick angle stiffeners, were used to validate the
tables. These angle stiffeners no longer predominate. Stronger steel is now used in ship
construction and the stiffener design has changed: ships are increasingly constructed with
bulb stiffeners. Figure 4 below is a photograph showing, side by side, both a 4 x 3 x
angle stiffener and a 240 mm x 10 mm (9.45 x 0.39) bulb stiffener. While the same
basic heat transfer and insulation issues confront us, the design is different and so the
assumptions, upon which the SNAME 4-7 design methodology is based, are no longer
valid. Because of this, we have a choice: either conduct a whole new set of heat loss tests
such as were done over 40 years ago, using the new bulb stiffener design, or have an
engineer, skilled in the use of a FEA heat transfer computer program, model these new
stiffener configurations and perform analyses to find solutions for particular design
problems.

Figure 4: A photograph of a 240 mm x 10 mm bulb stiffener currently used compared to a previously


common 4 x 3 x angle stiffener. Note that the larger bulb stiffener, while obviously stronger, would
also result in increased heat transfer from the ambient to the conditioned ship interior without additional
insulation.

Why there is a need for insulation


improvements
While a cursory comparison to DOE recommendations for home building insulation
thicknesses makes the SNAME 4-7 requirements look inadequate, an ocean going vessel
is not a house. We need to look in more depth at energy use on a ship to determine what
hull insulation actually does.

The space air heating energy on a ship is provided by electrical resistance heat. Diesel
generators generate the electricity with an operating efficiency of about 40% (i.e., the
ratio of electrical energy generated to the energy content of the diesel fuel burned). To
generate a kilowatt-hour of electrical energy then, one can compute the amount of diesel
fuel that must be burned. Assuming a diesel fuel content of 19,000 Btu / lb, calculations
show that about 1.2 lbs. of diesel fuel, or about 1/5th of a gallon, is needed to generate
each kilowatt-hour of electricity. For electric resistance space heating, we can safely
assume 100% conversion of electrical energy to heat.

For the purposes of an example, let us assume that we have insulated the above water
portion of the hull with the solution found above from the SNAME 4-7 Table 17: 2 of
hull board insulation over the hull and 1 of hull board over the angle stiffeners, an
insulation configuration which we saw provides a U value of 0.131. Let us further
assume that a ship travels for a month between refueling stops and that it experiences
heating season conditions with an average indoor to outdoor temperature difference of
40 F. Under these circumstances, calculations show during that month, each 1000
square foot of hull surface area would require about 180 gallons of diesel fuel. Assuming
a price of $1.12 per gallon for diesel fuel, the monthly fuel cost would be about $200 per
1000 square feet of hull for space heating. Is this expensive? That would depend on the
ship itself and on the duration of its operation between refueling stops.

Now, to determine whether it might be worthwhile to better insulate the hull, let us
propose using 4 of hull board insulation, instead of 2, but still with 1 on the angle
stiffeners. According to Table 19, page 94, of the SNAME 4-7 document, that insulation
system would provide a U value of 0.065, about half the 0.131 U value with 2 hull
board thickness and Inside Air to Weather Air, horizontal heat flow (see Attachments 1

and 2). Calculations then show that this 2 additional insulation thickness would of
course add weight, about 470 pounds of weight per 1000 square feet of surface area, but
would reduce the diesel fuel use by half, on a square foot basis. That is, the monthly use
of 1,370 pounds of diesel fuel consumed per 1000 square foot could be cut in half, to 680
pounds of diesel fuel per 1000 square foot, just by doubling the thickness of the
insulation board, only adding 470 pounds per 1000 square feet of thermal insulation!
That represents a monthly weight payback savings of 220 pounds per 1000 square feet
of surface area. Further, the monthly cost of the diesel fuel saved would be about $95 per
1000 square foot simply considering the value of the reduced use of diesel fuel. So, in
summary, the 2 of added insulation board would significantly reduce the weight and
reduce operating cost of the ship in just one month of operation in one of the earths more
extreme climatic zones.

It is important to note that the reduction in heating and cooling loads, resulting from the
use of thicker hull board insulation, could also reduce equipment weight and equipment
cost. That is, the size of the diesel generators and the size of the cooling units could
consequently be reduced. That, in turn, would reduce weight and reduce first cost of
constructing the ship, in spite of the thicker hull board insulation.

It is beyond the scope of this paper to perform detailed calculations for weight reductions
achievable by using smaller diesel generators. However, as an example, a 21,600 kW
marine style diesel generator from a particular manufacturer is listed on his Web site as
having a weight of 264 metric tons. If that same manufacturers 14,400 kW diesel
generator, which is listed as having a weight of 190 metric tons, could replace this larger
one, then the overall ship weight could be reduced by 74 metric tons. This is a significant
weight reduction.

FEA Modeling As A Tool


Structural finite element analysis (FEA) modeling is a design tool increasingly used by
the shipbuilding industry to analyze steel stress and fatigue through similar steel mesh
modeling techniques as used to analyze heat flows. Naval architects can design and test
their structural models against simulated extreme sea conditions with the aid of
computers. Similarly, and with similar levels of confidence, insulation designers can
design and test expected heat flows, using thermal FEA models, given input
environmental and ship conditions to arrive at proper insulation thicknesses. Given the
connectivity between heat flow and structural design, the insulation designer must

insulate what the naval architect designs. He can do this using FEA modeling. It is a
powerful tool that can greatly assist both the structural designer and the insulation
designer in the determination of proper ship design.

Regardless of the maximum U value requirements in Table 2 of the SNAME 4-7


document, the US ship building industry, the US Navy, and other owners and
constructors of ships could also benefit from FEA modeling of heat loss or gain through
the insulated hulls, bulkheads, decks, and other surfaces with the heavier scantlings. The
benefit of doing this would be to obtain more accurate and representative results. FEA
modeling provides more accurate overall material and surface temperatures from which
system U values can be calculated. Figure 5 below shows the results of one FEA on a
bulb stiffener surrounded with hull board (for a 70 F indoor temperature and a 13 F
outside weather temperature):

Figure 5: FEA temperature profile for a 370 mm bulb stiffener with 3 inches of hull
board on the plane surface and 2 inches of insulation on bulb under 70 F indoor air,
winter conditions with 13 F outdoor air, and horizontal heat transfer.

The advantage of FEA modeling, such as was done for the above case in Figure 5, is that
different insulation thicknesses, different bulb sizes and spacings, and different indoor
and ambient conditions can be evaluated with essentially the same model. This makes
FEA modeling an extremely powerful design tool. It also allows us to derive a more
accurate U value for a particular ship bulkhead construction.

Conclusions and Recommendations


For over 40 years, the US shipbuilding industry has relied on, or been influenced by, the
same standard for designing insulation systems for ships structural boundaries,
bulkheads, and decks. Shipbuilding and design engineers have relied on a document,
SNAME Technical & Research Bulletin 4-7: Thermal Insulation Report, for this
standard. While this document represents excellent research for its time, insulation
practices in the United States have made major advancements in those 40 years, thereby
leaving the SNAME 4-7 U values inadequate for todays thermal insulation guidelines.
Further, given the shipbuilding industrys concern for reducing ship weight, the
inadequate thermal requirements result in excess weight and cost of consumption for
diesel fuel, as well as excessive costs for over-sized diesel generators, and for over-sized
mechanical HVAC equipment. Finally, in recent years, the US shipbuilding industry has
changed to more robust scantling practices and consequently changed the predominant
stiffener design from angle type to a bulb type. This brings application of the results and
recommendations in SNAME into serious question.

These authors strongly recommend that the US shipbuilding industry undertake a new
project to evaluate and revise the maximum allowable U values for hulls, bulkheads,
and decks. This undertaking should include considerations for minimization of overall
ship weight, not just insulation weight, allowing tradeoffs between insulation weight
increases on the one hand and weight savings for fuel and mechanical equipment on the
other hand. The development of new U values should also include considerations for
minimization of fuel cost and labor to install thermal insulation. Finally, these authors
recommend the writing of a specification for marine insulation that encourages the use of
FEA to evaluate different insulation strategies.

References:

1. Technical & Research Bulletin 4-7, Thermal Insulation Report, The Society of
Naval Architects and Marine Engineers, 1963.
2. United States Department of Energy Insulation Fact Sheet, R-Value
Recommendations, available on the Internet through the Web address:
http://www.ornl.gov/sci/roofs+walls/insulation/ins_16.html

Attachments:

1. A copy of page 76, Table 17 with 2 of insulation, from Reference 1.


2. A copy of page 94, Table 19, with 4 of insulation, from Reference 1.

3.

You might also like