Beruflich Dokumente
Kultur Dokumente
377-386
Cargo vessels carrying heavy liquid petroleum products, such as No. 6 fuel oil or asphalt, traditionally have
an extensive heating system. This system consists of thousands of feet of 11/2 or 2-in. pipe arranged in the
bottom of the cargo tanks through which a heat-transfer fluid is passed. A heater is used to raise the tem-
perature of the heating fluid, which is forced through the coils. This maintains the cargo at a temperature
suitable for off-loading by centrifugal cargo pumps. This heat source (heater) for the fluid is fuel fired usually
with diesel oil. With the costs of fuel spiraling upward, an approach that would reduce fuel consumption
should be attractive to the owner/operator of such a vessel. This paper presents the results of an analysis
to determine potential savings in heating systems for No. 6 fuel oil cargos. Since positive-displacement
pumps can handle much more viscous fluids than centrifugal pumps, they were studied for off-loading in-
stead of using traditional vertical deep-well mixed-flow pumps. Two types of petroleum cargo vessels were
analyzed, including an oceangoing 180 000-bbl Tank Vessel and a river tow consisting of eight tank barges
with a combined capacity of 160 000 bbl. The analysis indicates that a significant reduction in fuel con-
sumption could be realized with the new cargo pump type by reducing the size of the heaters. Additionally,
the quantity of heating coils could be reduced. This would reduce construction costs and save substantial
operating costs if incorporated into petroleum cargo vessel design.
Background
+I
OUT BOARD COIL
J'~
THERE ARE large numbers (over 4000) of unmanned tank
vessels carrying petroleum and chemical products on the inland
waterways systems of the United States. There is also an in-
creasing number of oceangoing manned and unmanned vessels
carrying petroleum products in the Gulf and ocean trade f RETURN HEADER
routes.
River barges are operated in groups (referred to as "tows")
consisting of four to eight tank barges pushed by a single towboat. SUPPLYHEADER
Ocean barges are usually operated singly and are either pushed mu. D,L C,.CD.A..U .umP
or towed by a tug. Tankers are self-propelled and have very large
cargo capacities compared with barges.
,.OOARD
The cargoes comprise any liquid industrial product, including
the full spectrum of petroleum products. For light products (for
example, gasoline, jet fuel, diesel oil and heating oil) vertical
deep-well (VDW) pumps (driven by diesels) are established as
Fig. 1 Typical tank vessel heating system
a common method of discharging the cargo from barges. These
pumps are used for any liquid product with viscosities up to 1500
SSU (Saybolt Seconds Universal). Above 1500 SSU the efficiency heater fired with No. 2 diesel oil, a diesel generator set to provide
of a VDW p u m p drops rapidly and priming the pump becomes electrical power to the heater, and an extensive piping system
a serious problem. with supply and return headers and coils in each tank to circulate
In order to compensate for this performance ceiling, a heating the heat-transfer fluid. The heating system uses a special heat-
system is installed on the barge for use when high-viscosity transfer fluid (commonly called hot oil) and a pump to circulate
cargoes, such as oil or asphalt, are to be transported. These sys- the fluid through the coils to heat the cargo. Figure 1 is a sim-
tems heat the cargo to an acceptable viscosity (high temperature) plified diagram of a typical "hot oil" heating system. Figure 2 is
prior to starting the pump-off operation. This heating operation a typical oil-fired heater. Reference [1] 2 includes an excellent
in cold weather can continue for several days and is generally discussion of these systems and the problems and difficulties in
done en route so that when the barge arrives at its destination, sizing, selecting and operating them.
it can immediately start off-loading. The temperature generally For products heavier than No. 6 oil, such as asphalt, positive-
required for No. 6 oil prior to starting pump-off operations is in displacement (PD) pumps mounted on deck are used. They re-
the range of 120 to 150F. quire a suction strainer for protection and depend on their suction
The heating system for the barges consists of a barge-mounted lift capabilities and proper design of the suction piping to dis-
charge the product. There have even been situations where a PD
z General manager, Southeastern Division, Advanced Marine Enter- p u m p and a VDW pump are mounted on the same barge using
prises, Inc., Virginia Beach, Virginia. a common diesel engine driver. The pump selected for discharge
Presented at the October 1, 1981 meeting of the Hampton Roads is based on the cargo (for example, PD for asphalt and VDW for
Section of THE SOCIETY OF NAVAL ARCHITECTS AND MARINE EN- gasoline). Vertical deep-well pumps have been used for asphalt,
GINEERS.
Author's note: Certain equipment manufacturers are mentioned by but this is not recommended.
name in this paper. This is not to be construed as an endorsement of their T y p i c a l v e r t i c a l d e e p - w e l l p u m p . This p u m p consists of a
products by the author. There are other manufacturers that provide such
products and purchase selection should be made on technical merit. 2 Numbers in brackets designate References at end of paper.
_L
Fig. 2 Typical oil-fired heater
BOTTOMPL ~ )'
series of impellers mounted in a well with the priming stage as
close as practical to the bottom of the well. The impellers are
driven with a shaft running to the discharge head mounted above Fig. 3 Typical vertical deep-well pump
deck. A right-angle gear drive is mounted on the discharge head
and the driver (typically a Detroit Diesel 6-71 or 8V-71) is con-
nected to the gearbox. Figure 3 illustrates a typical multistage mounted or mounted on a vertical column as shown in Fig. 5.
barge pump. A typical pump operating at 1770 rpm while PD pumps can handle a wide range of viscosities; the only limit
pumping a fluid of 1500-SSU viscosity is rated at 2100 gpm @ is available horsepower. A GTS pump operating at a speed of
100 psi and requires 245 hp. 1770 rpm is rated at 2031 gpm.
Positive-displacement pumps. Positive-displacement pumps
are available in a number of types, including screw, gear and vane. Comparison of VDW and GTS pumps
The volume of liquid moved is a direct function of the speed of
the pump. The pump selected for this study is a geared twin- Comparing VDW and GTS pumps is essentially comparing a
screw (GTS) as illustrated in Fig. 4. These pumps may be deck centrifugal pump with a positive-displacement pump. Centrifugal
POWER U~ DRIV.......
RIVE
PIPING
JABOVE
LINE!
MAINDECK
4.
DISCF
PIPING
S,L
~EAFIS
/'-- BOTTOMPLA ,,
Fig. 4 Typical construction of a geared twin-screw pump Fig. 5 GT$ vertical cargo pump wet-well installation
l
I I SPEED: 1770 RPM /
L Iv'soos'TY: ,500sso
90
80
I MODEL:
HEAO
GTS 208-112
SPEED: 1770 RPM
t 70 V SCOSTY: 5000 SSU
1,[.oo
300 BNP ~ JOTOOOSSU
r .P o Sr0OOSSO
/%/ "'" 2DD'; B.. O..GOOGGD
t'SO [
tIDGg 'ot
50 ~ 20
pumps have no inherent suction lift or self-priming features; they Fig. 7 Performance curve for GTS pump
have to be designed and added in to the system. GTS pumps have
both of those characteristics inherent in their design and can single-skin barge gives the owner maximum cargo-carrying ca-
provide a suction lift up to 27 in. Hg (30 ft of water). Centrifugal pacity for given length, width and depth. It provides the most
pumps can pass some solid material without damage. PD pumps capacity for the least amount of steel and is the simplest barge
require suction strainers of some type since small hard objects to construct. Thus, it is the least expensive barge to purchase;
could harm the rotors. VDW pumps have to be mounted with the however, the single-skin barge is easiest sunk and has the greatest
priming stage as low as possible in the barge. GTS pumps could chance of cargo loss if it is in a high-energy collision. It also allows
be mounted on the deck of the barge, if the available suction head the greatest loss of heat to the environment (water and air) of the
(vapor pressure, static lift and friction losses combined) is not three types.
exceeded. This is normally not a problem on river barges. Double-wall, single-bottom barges are a compromise type that
VDW pumps are severely limited by higher viscosities and will better withstand high-energy collisions with other barges and
operation above 1500 SSU reduces efficiency significantly. GTS piers, but are still vulnerable to grounding damage.
p u m p efficiency is relatively constant from 500 to 10 000-SSU Double-hull barges utilize a full double bottom and wing void
viscosity and a GTS pump will maintain the same flow through construction which provides maximum protection in all types of
this range, within 5 percent. accidents. It is the most expensive type to construct and has the
The curves shown in Figs. 6 and 7 were developed to compare least amount of cargo capacity for the amount of steel. The U.S.
the two types of pumps operating at the same speed (1770 rpm). Coast Guard is pushing for regulations t h a t would limit con-
The VDW pump curve pertains to a viscosity of 1500 SSU struction to this type of barge only, b u t the industry is resisting
whereas the GTS pump curve is for 1500 to 10 000 SSU. This this effort [2]. The double-hull barge has the lowest heat loss to
viscosity difference compares to a liquid temperature difference the environment, since the double-hull construction has an in-
of approximately 46F. The important factor is the horsepower sulating effect.
required. Even though the GTS is pumping much more viscous
liquid, the additional horsepower required is minor. For 2100 gpm
at 100 psi, the VDW pump requires 245 hp; the GRIPSrequires 225
hp at 1500 SSU and 287 hp at 10 000 SSU. A Detroit Diesel Model
CARGO TANK CARGO TANK
8V-71 is rated at 304 hp for intermittent duty, thus would satisfy
all these horsepower requirements. Table 1 summarizes these
characteristics at the various viscosities.
SINGLE SKIN TYPE
l
Double wall, single bottom
Double hull CARGO TANK CARGO TANK
~ qEo~
ooFI 0 o~
q~ =-o 0 ~o-=
[]BALLAST ~] CARGOFUEL [~] VOID onds Universal for all discussion of viscosity. The viscosity range
covered herein is 1500 to 10 000 SSU. For comparison, water at
68F has a viscosity of 30 SSU.
The specific cargo selected for this study was No. 6 fuel oil,
since it is a fairly common cargo currently being carried in barges
and :requiring heating. The results could be extrapolated to other
types of cargoes requiring heating if viscosity characteristics are
BARGE
available. Figure 11 is a graph of the effect of temperature on No.
Fig. 10 Typical single-skin ocean tank barge 6 oil for the range 10 to 150F. It should be noted that at a given
viscosity the corresponding temperature could vary by 25F ei-
Since barges are being constructed of all three types, this study ther way.
addresses the heating requirements for all three. A,~ described earlier in the discussion of the pumps, viscosity
Inland river tank barges. These barges are shallow draft greatly affects the performance of a VDW pump, with 1500 SSU
(typically 8 ft-6 in.) and are relatively flat bottom. They are being the upper limit. As the viscosity increases, the brake
moved in groups of four to eight called tows. They are restricted horsepower increases markedly, the head is reduced significantly,
to rivers and are not capable of ocean operation. Figure 9 illus- and there is some reduction in the capacity.
trates a typical tow of barges. Establishing heating requirements. The required conditions
Ocean tank barges. Ocean barges are pushed or towed by for a pump to discharge No. 6 cargo dictate the temperature to
tugboats, have drafts up to 35 ft and are designed for ocean op- which the cargo must be heated. For VDW pumps, this temper-
eration with a hull form similar to ocean tankers. Figure 10 il- ature ranges from 120F m i n i m u m to as high as 150F. In most
lustrates a typical ocean tank barge. cases it is a subjective judgment by the tankerman, who must
consider the characteristics of the particular equipment and cargo
Heating analysis involved.
Since very little actual data are available from the field, the
Viscosity considerations. Viscosity is that property of any appendix in reference [1] is heavily relied upon for practical data.
fluid which tends to resist a shearing force. The viscosity of a A temperature of 144F was chosen as the temperature that a
liquid is temperature dependent. This paper uses Saybolt Sec- cargo is typically heated to prior to starting pump-off. This cor-
responds to a viscosity of 1500 SSU, which is at the upper end of
TEMPERATURE(F) No. 6 oil characteristics in Fig. 11. It is also the upper performance
limit of VDW pumps.
0 20 40 60 80 I00 120 140
For a typical cargo to be pumped by the GTS pump, 100F was
I00,000 chosen as the temperature required for pump-off. This could have
50,000 been much lower, but was used since it corresponds to approxi-
mately 10 000 SSU.
20,000 In order to establish how much heat is required, the temper-
ature of the cargo prior to starting the heating system must be
I0000 established. Since the cargoes are generally loaded hot, it was
considered reasonable to let the cargo cool to 60F prior to
starting the heating system. This is lower than is normally done
5000 in practice, but since the GTS will pump 100F cargo with t h e
same driver as the VDW pump, it is considered a reasonable lower
temperature bound. Additionally, based on the a m o u n t of heat
required, it would be more efficient to run the heating system for
1500
a shorter time even though the temperature difference is higher.
This corresponds to starting the heater on the fifth day of a
>
ten-day trip.
500 A m b i e n t temperatures used for these calculations were se-
lected as 40F water and 28F air. If lower temperatures had been
selected, the heating requirements would increase. A worse case
occasionally encountered is 32F water and air temperatures
below 0F. For purposes of this study, these temperatures were
considered too extreme to consider for comparison purposes.
In order to determine an average temperature to calculate an
overall heat-transfer coefficient, it was assumed that 70 percent
of tile vessel was exposed to water and 30 percent exposed to air.
100 A change in this ratio would affect the calculations only slightly.
Fig. 11 Effect of temperature on viscosity A change in the ambient temperatures selected would have a
AIR
CONVECTION
\ LJ CONVEC~ON &
CONVECTION RADIATION
SECTION
AIR
~NVECTION
WATER RADIATION
CONVECTION
ELEVATION
Fig. 12 Heat losses to the environment
much more significant effect on the heating requirements. and air during transient. The sum of these two types of heat is
The heat transfer is dependent on whether the vessel is moving the amount of heat required to be added to the cargo so that its
through the water or tied up at a pier. To determine heat loss, it discharge temperature requirement is met (144F for the VDW
was assumed that the vessel was always moving. This is a rea- and 100F for the GTS).
sonable assumption, since pumping commences as soon as the Sensible heat is a direct function of the volume of cargo and
vessel is moored. could be expressed as average heat per unit volume per degree
(147 Btu/bbl-F) for extrapolating this information to other size
H e a t n e c e s s a r y to m e e t r e q u i r e m e n t s vessels.
Heat loss to the environment is a function of the boundaries
Two types of heating are addressed in this paper:
Sensible heat of the vessel and is affected by the type of construction, tem-
Heat loss to the environment peratures selected, and velocity through the medium. Extrapo-
Sensible heat is the heat required to raise the cargo from its base lating this information to vessels other than those considered
temperature to the temperature required for pump-off. The herein would give reasonable results if the relative sizes are
sensible heat required is calculated using the equation: properly considered. For vessels larger than those in the study,
the extrapolation would yield fair results if the sensible heat was
Q = CppAT (5.6) Cbb1 used as a direct ratio of volume and the heat loss was a ratio of
where: the areas of the vessel boundaries. Figure 12 illustrates the heat
loss to the environment.
Q = sensible heat For the three types of vessels considered, Table 2 summarizes
Cp = specific heat the heating loads for sensible heat, heat loss, and average heat
p = density (Btu/hr) required for a five-day heating cycle. This is to meet the
A T = temperature difference VDW pump requirement of 144F cargo temperature and the
5.6 = factor (cubic feet per barrel) GTS pump requirement of 100F before the starting pump-off
Cbbl = capacity in barrels operations. The Appendix gives the procedure for calculating
those values.
Loss to the environment is the amount of heat lost to the water It must be emphasized that the GTS pump can handle much
ENVIRONMENTALCONDITIONS TERMINOLOGY
ambient air temperature = 28 F double-wall, single-bottom vessel--22000-bbl capacity (DW/SB)
ambient water temperature = 40 F double-hull vessel--20 000-bbl capacity (DH)
initial cargo temperature = 60 F single-skin vessel--180 000-bbl capacity ,(SS)
cargo temperature for GTS = 100 F geared twin-screw pump (GTS) pump
cargo temperature for VDW = 144 F vertical deep-well pump (VDW) pump
Vessel Type
DW/SB DH SS
GTS VDW GTS VDW GTS VDW
Linear feet of 2-in. pipe required 6300 12 500 6300 12 500 60 000 120 000
Difference of 2-in. pipe required (VDW less GTS) 6200 6200 60 000
Estimated $ cost of difference 31 000 31 000 300 000
ENVIRONMENTALCONDITIONS TERMINOLOGY
ambient air temperature = 28 F double-wall, single-bottom vessel--22 000-bbl capacity (DW/SB)
ambient water temperature = 40 F double-hull vessel--20 000-bbl capacity (DH)
initial cargo temperature = 60 F single-skin vessel--180 000-bbl capacity (SS)
cargo temperature for GTS = 100 F geared twin-screw pump (GTS) pump
cargo temperature for VDW = 144 F vertical deep-well pump (VDW) pump
Vessel Type
DW/SB DH SS
GTS VDW GTS VDW GTS VDW
Total gallons of oil for 5-day trip 2760 5640 2760 5640 27 120 54 240
Difference (VDW less GTS) 2880 2880 27 120
Since the cost of diesel fuel is rising rapidly, costs were not For assistance in preparing the paper, I wish to thank Mrs.
determined. It would be straightforward to multiply these figures Doreen Davis for the typing and Miss Renee Norman for the
by current/projected price per gallon to determine savings. graphic artwork.
Summary References
This study determined that if geared twin-screw pumps were 1 Gundlach, James O., "Heating Cargoes in River Barges," SNAME
installed for cargo off-loading in lieu of vertical deep-well pumps, Spring Meeting, New Orleans, 1974.
the heating requirements for petroleum vessels would be sub- 2 Lawrence, Berdon, "Double Hulled Disaster," The Work Boat, Jan.
stantially reduced for No. 6 fuel oil cargoes. 1980, p. 73.
3 Gartmann, Hans, Delaval Engineering Handbook. 3rd ed.,
The savings in an initial installation for a single-skin ocean McGraw-Hill, New York, 1970.
vessel would be very significant. For a double-hull river vessel, 4 Chapman, Alan J., Heat Transfer, MacMillan, New York, 1960.
the initial installation savings would be small. This information 5 Hodgman, Charles D. et al, Handbook of Chemistry and Physics,
is summarized in Table 6. Chemical Rubber, Cleveland, Ohio.
6 Hicks, Tyler G. and Edwards, Theodore W., Pump Application
The savings in fuel consumption would be very significant in Engineering, McGraw-Hill, New York, 1971.
all types of vessels considered, with the single-skin vessel the 7 Baumeister, Theodore and Marks, Lionel S., Standard Handbook
highest. for Mechanical Engineers, 7th ed., McGraw-Hill, New York, 1967.
For a new design vessel, an analysis should be done of the
heating requirements using anticipated environmental factors, Discussers
From a heat loss standpoint the following items would be con-
sidered: Charles J. Hornbostel Horace Steven
Use double hull construction. W. D. Burton, Jr. J. E. Ancarrow
Install GTS cargo off-loading pumps.
Insulate the main deck.
Table 6 Summary of initial installation savings
Install the hot-oil supply and return headers below the main
deck.
Approximate Savings (thousands of $)
Reduce the quantity of heating coils by performing a suitable
heating analysis. SS Vessel DH Vessel
Item 180 000 bbl 20 000 bbl
Incorporating these items in the design of a vessel will raise the
initial construction cost; however, the substantial savings in Heaters 180 10
operating costs due to reduced fuel consumption should easily Generators 50 nil
justify the initial cost. Piping 300 31
Pump driver nil nil
Determination of heat loss factor for typical Determination of heat loss factor for single-skin
double-wall/single-bottom barge ocean tank vessel
Barge data (see Fig. 14): The Appendix of reference [1] provides the basic heat-transfer
equations for heat loss from typical barges. These equations are
length = 295 ft width = 54 ft depth = 13.5 ft well known and discussed in all heat-transfer texts. (for example,
cargo = No. 6 oil capacity = 22 000 bbl Reference [4]).
loading temperature = 150F Determining the relevant properties of No. 6 fuel oil and sea-
unloading temperature = 137F water at several temperatures is required prior to initiating the
water temperature = 50F calculations. Data for these properties were obtained from ref-
air temperature = 45 F erences [3,5,6].
Since the vessel under consideration is a single-skin barge with
shell thickness of 3/s in., the thermal conductivity of the skin is
A m b i e n t t e m p e r a t u r e (TA):
quite high.
TA = 0.7 TH20 + 0.3 Tair A fouling factor of h = 1000 was applied to the surface of the
barge skin exposed to the ocean, which includes paint and sea
TA = 0.7(50F) + 0.3 (45F) growths per reference [7], which yields lower heat-transfer rates
than clean steel.
TA = 48.5F The d a t a developed and used in this determination are listed
Solve for h e a t loss c o n s t a n t (UoA): in Table 7.
No steady-state condition appears on Fig. 14 [1] so average The determination of heat loss to the environment first re-
values for cargo temperature and heat added (Btu/hr) must be quired determining the film heat-transfer coefficients to water
, ,
120
,x,,,, I
COMIJTER CJRVEj ,
I--
_J
+
0 20 40 60
80 I00 120 140 160 180 200 220 240 HOURS
! ! I I I I i I I I I
0 1 2 3 4 5 6 7 8 9 10 DAYS
TIME FROMLOADING(HOURS& DAYS)
Fig. 14 Comparison of predicted with test data (Chart 4A from Reference [ 1])
140 /
BTU/IR ADDED
130 ~ NO HEAT )DED----~.~. xlOb,,2. _5,~.~, .Oxl(
XIUO I
120
\ I I
I i
o
v
I
110
,
20. or -,
BOTTO l \
~n lO0 ,,,,.,,,
65~0 FT
70
I----2" S(H 4 0 - - ~ - - , , ,
PIP
60
50
? 201 40 1 60 1 80 100 1~0 14V 160 1 180 1200 220 210 HOURS
0 i 2 3 ~ 5 6 7 8 ~ 10 DAYS
TIME FROMLOADING (HOURS& DAYS)
Fig. 15 Comparison of predicted with test data (Chart 4B from Reference [1])
Oil Properties
T~ F p lbm/ft ;~ ~ 1/F Cp Btu/lbm-F h Btu/hr-F-ft t~ lb/ft-hr
Water Properties at 40 F
p ~ Cp k tt
lbm/ft 3 1 F Btu/lbm_OF Btu/hr-F-ft lb/ft-hr
or Uo = overall h e a t - t r a n s f e r c o e f f i c i e n t
A = area of h e a t t r a n s f e r
where C = 0.13 w h e n NGR" NpR > 109 AT = temperature difference
a = 0.33
Metric Conversion Table
NOTE: T h e values of t h e c o n s t a n t s c, a, a n d b were d e r i v e d Original experimental data herein were measured in U.S. customary
f r o m r e f e r e n c e [1]. units. Following is a list of standard factors fbr conversion to metric (SI)
T o d e t e r m i n e h/ for s e a w a t e r on o u t s i d e of barge, walls or values:
bottom: 1 ft = 0.3048 m
1 ft 2 = 0.092903 m 2
1 ft 3 = 0.028317 m :~
k (N RE)a(NpR) b
hi = C-~ 1 in. = 25.4 mm
1 psi = 6.894757 kPa
1 mile = 1.6 km
w h e r e C = 0.664 w h e n NE < 400 000 a n d Npl~ > 0.6 1 hp = 0.7457 kW
1 gal (U.S.) = 3.785 412 liters
a = 0.5 1 bbl = 0.158 987 m :~
b = 0.33 C = '~'/9(F-32)