You are on page 1of 6

Ultra low sulphur diesel fuel

impacts on availability of other


fuels
G. R. Hadder, S. Das, R. Lee, N. Domingo and R. M. Davis

Legislation for ultra low sulphur (ULS) diesel particulate limitations would require the use of traps
fuel, with a greatly reduced allowable sulphur on the vehicle exhaust system. The industry express-
content and a new limit on aromatics content, is ed concern that these traps might be ineffective with
expected to be in place in the USA by 1995. The present day sulphur levels in diesel fuel because of
ULS diesel fuel has been specified to satisfy US the trap plugging by sulphates in the engine exhaust
standards for particulate emissions from heavy- and the generation of increased sulphur emissions
duty diesel engines. The Navy Mobility Fuels catalyzed by the traps. Commenters recommended
Forecasting System has been used to study the that EPA reduce the allowable sulphur in diesel fuel.
impacts of ULS diesel fuel production on other EPA responded that it would continue to study the
refined petroleum products. The study predicts issue and, if warranted, would consider so doing. 3
that to produce ULS diesel fuel, US Gulf and While addressing the issue of diesel fuel sulphur,
west coast refiners will have to invest about $4.4 EPA investigated the emission effects of other diesel
billion in new processing capacity. Refiners will fuel properties such as aromatics content. 4 Aromatic
shift aromatics from No 2 diesel fuel to jet fuel compounds in fuel contribute significantly to both
and to No 2 fuel oil. Therefore, particulate the carbonaceous fraction of particulate emissions as
emissions could be transferred from highways to well as hydrocarbon emissions from heavy-duty
airways and to communities which use No 2 fuel diesel engines under transient conditions. 5
oil for residential and commercial heating. The Various studies were conducted to estimate the
study also predicts that there will be an increase cost to cut diesel fuel sulphur content to 0.05 weight
in the aromatics content of Navy JP-5 jet fuel and percent 6 and to limit diesel aromatics content to a
F-76 marine diesel fuel produced in the USA. maximum of 20 volume percent. 7 EPA now esti-
The gum-forming tendencies of F-76 and F-77 mates that the cost to produce this ultra low sulphur
burner fuel oil will increase in most cases. The (ULS) diesel fuel will be equivalent to 7 c/g of diesel
freezing point of JP-5 will improve. There will fuel, with an industry investment of $2.7 billion, s
be minor changes in the cost of JP-5, but sizable Based on an independent survey, the petroleum
reductions in the cost of F-76 and F-77. refining industry has estimated that the cost to
produce ULS diesel fuel will be equivalent to 15.5
Keywords:Diesel fuel; Jet fuel; Environmentalregulations c/g, with an industry investment of $6.7 billion. 9
Industry analysts believe that legislation covering
the levels of diesel sulphur and aromatics is likely to
On 15 March 1985, the USA Environmental Protec- be in place by 1995.1°
tion Agency (EPA) established particulate standards A plausible impact of the production of ULS
for heavy-duty diesel engines. 1 These standards diesel fuel would be that some of the high sulphur
were to go into effect for 1988 model year vehicles, and high aromatics stocks currently blended to diesel
and were to follow with progressively more stringent fuel would be diverted to products with relatively
regulations. 2 In response to these regulations, the high allowable levels of sulphur and aromatics. For
diesel engine manufacturing industry stated that the example, F-76 marine diesel might be a good dis-
position for high sulphur stocks, since the limit on
The authors are members of the Energy Division, Oak sulphur in F-76 is 1.0 weight percent or 20 times the
Ridge National Laboratory, PO Box 2008, Oak Ridge, limit for ULS diesel fuel. The purpose of this report
Tenessee, TN 37831, USA. is to discuss the application of the Navy Mobility

488 0301-4215/91/050488-06 0 1991 Butterworth-Heinemann Ltd


Ultra low sulphur diesel fuel impacts on other fuels

Table 1. Key specifications for navy fuels.

JP-5 F-76 F-77


API gravity 36-48 11.5 minimum
Flash point, °F, minimum 140 140 150
Freezing point, °F, maximum -51
Smoke point, mm, minimum 19
Cetane number, minimum 45
Sulphur, wt %, maximum 0.4 1.0 3.5
Pour point, °F, maximum 20 15
Particulate matter, mg/L, maximum 1.0
Accelerated stability, mg/100 ml, maximum 1.5

Fuels Forecasting System in the assessment of such models: the oil market simulation model; the pet-
impacts of ULS diesel fuel production on the availa- roleum allocation model; and the refinery yield
bility and quality of other fuels, with emphasis on model.
Navy jet and marine fuels.
The oil market simulation model
The econometric oil market simulation (OMS)
Model analysis of domestic production of model 11 estimates the price at which global supply
fuels and demand for oil are in balance. Aggregate sup-
The Navy Mobility Fuels Forecasting System has ply, demand and price data from the OMS model are
been used to predict the impacts of the production of used for some of the input data requirements of the
ULS diesel fuel on the production of other refined other two models of the forecasting system.
petroleum products. The study focuses on the effects The methodology of the OMS model consists of a
of ULS diesel fuel production on the availability and specific set of behavioural rules for three groups of
quality of Navy JP-5 jet fuel, F-76 marine diesel fuel, market participants who have different objectives:
and F-77 burner fuel oil. Key specifications for these oil consumers, OPEC oil producers, and non-OPEC
fuels are summarized in Table 1. oil producers.
The forecasting system, shown conceptually in The oil demand equations for the six regions of
Figure 1, consists of three Department of Energy the consumer group are expressed in terms of four

FORECASTINGSYSTEM

I
CRUDE ? I
AND SYNCRUDE I
CHARACTERISTICS .L . . . .
l- . . . .
I SUPPLYDISRUPTION
"1 SCENARIOS
L
WORLD
CONVERSION DISTRIBUTION
TECHNOLOGY CRUDE,PRODUCT
TRENDS (PAL)

REGIONAL
.EF,NERY I REFINERY
TRENDS ] MODELLING

-I (REMS/RYM)

±
T
REFINEDPRODUCT(FUELS)QUANTITY,
QUALITY,COST

Figure 1. Navy Mobility Fuels Forecasting System and supporting activities.


Source: S. Das et al, Navy Mobility Fuels Forecasting System Phase IV Report,
ORNL-6517, Oak Ridge National Laboratory, Oak Ridge, TN, USA, 1988.

E N E R G Y P O L I C Y June 1991 489


Ultra low sulphur dieselfuel impacts on other fuels
variables: the world oil price; regional income; the Each regional RYM determines the optimal solu-
economic feedback effect of world oil price on tion which maximizes the gross refinery margin.
regional income; and regional oil demand in the Because of the detailed refinery technology repre-
previous year. sented within RYM, the model can be used to
For the group of OPEC oil producers, the OMS perform a host of impact analyses by estimating the
model incorporates a theory of pricing behaviour changes in product yields given changes in input
that maintains that the higher the demand for OPEC variables. These variables include capacity, utiliza-
oil relative to the OPEC maximum sustainable oil tion and yield structure for more than 20 refinery
production capacity, the greater will be the increase processes; refinery slates which can contain over 30
in the world oil price. The pricing theory, repre- products; product prices; detailed blending speci-
sented mathematically in a price reaction function, fications for most products; availability and cost of
assumes that the OPEC pricing objective is to main- more than 100 crude oils; crude oil assays; and cost
tain a target percentage of OPEC oil production of utilities.
capacity utilization. If the percentage falls below the RYM is idealized in that it assumes 'connected
target level, OPEC reduces the world oil price to refinery pipelines' within a large region. Conse-
encourage increased demand for OPEC oil and quently, RYM has a tendency to over-optimize
movement toward the target level. Conversely, refinery operations. In a verification study to deter-
when the actual capacity utilization rises above the mine how closely RYM could simulate refinery
target level, OPEC raises the world oil price in order activities in 1982, it was estimated that regional
to reduce demand and utilization of oil production refinery aggregation could have caused the model to
capacity. overestimate light product yields by about 3% of
The supplies of oil from the five regions in the total gasoline and distillate yields. 14 The over-
non-OPEC producer group are determined by the optimization problem can be mitigated by focusing
world oil price and the regional oil supplies in the on scenario-induced changes in refining operations,
previous year. The required relationship in an OMS rather than relying on the model to predict the exact
model solution is that the free world quantity of oil values of refining variables for a particular scenario.
demanded equals the quantity supplied. The OMS and PAL models produce an aggre-
gated picture of regional refining under different
The petroleum allocation model market scenarios. Although regional refinery yield
The petroleum allocation (PAL) model 12 is a linear models are embedded within the PAL model, they
programme that estimates global petroleum move- do not explicitly represent the production of impor-
ments for different market scenarios. Crude oil tant military fuels such as JP-5. Therefore, RYM,
supplies and the production of refined products, which contains enhanced descriptions of military
such as motor gasoline, are forecast by the PAL fuel production, uses some key results of the PAL
model for input to the refinery yield model. model but is executed independently of it. The
Two basic assumptions underlie the PAL model stand-alone RYM can be used efficiently to evaluate
objective function. First, the distribution of crude oil many variations of the refining environment. For
from supply points to refinery regions, and the example, RYM is used to estimate investment in
distribution of refined products from refinery re- domestic refining capacity and to produce detailed
gions to refined product demand regions will tend to reports on the effects of changing refined product
follow historical movements. Second, when supply prices and specifications.
conditions are such that these historical movements Given the global petroleum market analyses of the
cannot be satisfied, crude and refined products will OMS and PAL models, RYM has been used to
be distributed according to the pattern that mini- analyse the impact of ULS diesel fuel production. A
mizes total costs. Within the PAL model, there are region was selected for study if it had a recent history
33 crude oil supply regions; over 50 crude oil types; of production of Navy mobility fuels. The selected
37 refinery regions; 33 refined product demand regions were:
regions; and modes of petroleum transportation • US Bureau of Mines (BOM) Districts 7 and 11
including tanker, pipeline, rail and truck. (Inland Texas and New Mexico);
• BOM District 8 (Texas Gulf Coast, which
The refinery yield model contains over 85% of Texas's crude oil distilla-
The refinery yield model (RYM) is a linear tion capacity);
programme 13 that represents regional refining op- • BOM Districts 9 and 10 (Louisiana, Mississip-
erations. pi, Alabama and Arkansas); and

490 ENERGY POLICY June 1991


Ultra low sulphur diesel fuel impacts on other fuels
• BOM District 13 (Alaska, California, Hawaii, invest about $4.4 billion, according to the Navy
Nevada, Oregon and Washington). Mobility Fuels Forecasting System. is The invest-
ment will be primarily for the processes of distillate
Scenarios for production of current-quality diesel
deep hydrotreating, naphtha hydrotreating and gas
and ULS diesel were studied for petroleum markets
oil hydrocracking.
in the year 1995. In both scenarios, it was assumed
The production of ULS diesel fuel results in a shift
that motor gasoline will be unleaded and that the
of aromatics to civilian jet fuel and to No 2 fuel oil as
Reid vapour pressure of summertime motor gasoline
summarized in Table 2.
will be reduced by about 25% relative to the present
Table 3 shows some of the key effects of ULS
motor gasoline. 15
diesel fuel on JP-5:
Business-as-usual market conditions, which were
estimated with the OMS and PAL models, are • ULS diesel production causes an increase in
described in detail by Das et al. 16 A map which the aromatics content of JP-5 in all study
summarizes the projected business-as-usual free- regions.
world distribution of petroleum is shown in Figure 2. • Increases in aromatics content would be ex-
pected to result in decreases in smoke point.
However, the only substantial decrease in
Impacts of ULS diesel fuel production smoke point is in BOMs 9 and 10.
Of the nation's total crude oil distillation capacity • The forecasting system predicts substantial im-
66% is located in the Gulf and west coast study provements in the freezing point of JP-5 in
regions of BOMs 7, 8, 9, 10, 11 and 13.17 To produce BOMs 7, 8, 11 and 13. The improvement of
ULS No 2 diesel fuel rather than current-quality freezing point is consistent with the fact that
diesel, refiners in these study regions will have to freezing point improves substantially as the

Figure 2. Business-as-usual free-world petroleum (crude oil and refined product) distribution in 1995.
Source: S. Das et al, Navy Mobility Fuels Forecasting System Phase IV Report, ORNL-6517,Oak RidgeNationalLaboratory,Oak Ridge,
TN, USA, September 1988.

ENERGY POLICY June 199! 491


Ultra low sulphur diesel fuel impacts on other fuels

Table 2. Civilian product quality changes associated with change to ultra low sulphur diesel
specifications.

BOM 7/11 BOM 8 BOM 9/10 BOM 13


Jet A aromatics (vol %) +2.0 0 +3.6 0
No 2 diesel aromatics (vol %) -9.8 -13.8 -12.0 -17.9
No 2 diesel cetane index (number) +3.9 +8.3 +0.7 +3.6
No 2 diesel sulphur (wt %) -0.4 -0.4 -0.4 -0.3
No 2 fuel oil aromatics (vol %) +24.7 +25.9 +15.5 +16.2
No 2 fuel oil sulphur (wt %) 0 -0.2 0 -0.2

characterization factor of a fuel decreases (ie, Observations and recommendations


the fuel becomes more aromatic), for a con-
stant 50% recovery temperature. 19 The forecasting system predicts that the production
The change in cost of JP-5 ranges from -2.3 to of ULS diesel fuel will cause an increase in the
+1.4 c/g. aromatics content of domestically produced JP-5. It
is recommended that the impact of higher aromatics
Table 4 shows some of the key effects of ULS content in JP-5 be reviewed with respect to compara-
diesel fuel on F-76. bility with materials such as elastomers and with
respect to engine performance characteristics, such
• ULS diesel fuel production causes large in- as smoking.
creases in the aromatics content of F-76 in It is also predicted that the production of ULS
BOMs 8 and 13. diesel fuel will cause an improvement in the freezing
• In BOM 13, the cetane index of F-76 drops by point of domestically produced JP-5. It is recom-
9 numbers. mended that future research focus on the extension
• There are increases in light cycle oil blended to of JP-5 supplies, during market disruptions or mili-
F-76 in all regions of production of this fuel. tary mobilization, by blending higher freezing point
Light cycle oils have inferior stability, with stocks.
tendencies to produce gums and other particu- The production of ULS diesel fuel is predicted to
lates, resulting from complex interactions be- cause an increase in the aromatics content of domes-
tween molecular oxygen, organocompounds of tically produced F-76. It is recommended that the
nitrogen, sulphur and oxygen, and the more impact of higher F-76 aromatics content be reviewed
reactive hydrocarbon species such as olefins with respect to ignition and particulate formation
and aromatics. The resulting particulates and characteristics.
gums can cause plugging of diesel engine filters The forecasting system predicts that the produc-
and fouling of fuel injectors. 2° Studies have tion of ULS diesel fuel will cause an increase in the
shown rapid degradation in the stability of percentage of light cycle oil blended to domestically
fuels blended with more than 30% light cycle produced F-76 and an increase in the percentage of
oil in virgin distillate. 21 Hardy et al cite the light cycle oil blended to F-77 produced in BOM 13.
marked instability of a blend of 70% good It is recommended that because of potential fuel
quality F-76 with 30% light cycle oil. 22 stability problems, the possibility of high percen-
• The cost of F-76 is reduced in all study regions. tages of light cycle oil in F-76 and F-77 should be
discussed with refiners. These discussions could
The quality of F-77 produced in BOM 8 is not identify the need for follow-up investigations on the
affected by ULS diesel fuel production. However, in stability and power system impacts of such blends.
BOM 13 the sulphur content of F-77 increases Finally, it is predicted that the production of ULS
substantially and the change in light cycle oil diesel fuel will cause a shift of aromatics from No 2
blended to F-77 is +60 volume percent. In both diesel fuel to jet fuel and to No 2 fuel oil. Since
regions, the cost of F-77 is substantially reduced. aromatics tend to produce more smoke in the com-

Table 3. Key effects of ultra low sulphur diesel specifications on JP-5.

BOM 7/11 BOM 8 BOM 9/10 BOM 13


Aromatics content (vol % change) +1.6 +2.2 +1.2 +1.0
Smoke point (mm change) -0.5 +0.2 -5.2 -0.1
Freezing point (degrees F change) -25.0 -42.0 +2.0 -24.0
Cost (c/g change) +1.1 0 +1.4 -2.3

492 ENERGY POLICY June 1991


Ultra low sulphur diesel fuel impacts on other fuels
man shows a relationship in which a minimum cetane number of
Table 4. Key effects of ultra low sulphur diesel specifications on 40 (for a blend of light cycle oils and virgin distillates) corresponds
F-76. to a maximum aromatics content of about 38 volume percent. See
G. H. Unzelman, 'Higher diesel quality would constrict refining',
BOM 8 BOM 9/10 BOM 13 Oil & Gas Journal, 29 June 1987, pp 55-59.
Aromatics content (vol % change) +7.8 -1.2 +20.6 8Bob Williams, 'US refiners may face gasoline supply crunch in
Cetane index (number change) 0 0 -9.0 driving season', Oil & Gas Journal, 5 June 1989, pp 23-28
Light cycle oil (vol % change) +13.5 +10.0 +35.4 90p cit, Ref 6.
Cost (c/g change) -4.0 -2.4 -9.6 l°Op cit, Ref 7.
HSystem Science Inc, The Oil Market Simulation Model, Model
Documentation Report DOE/EIA/19656-2, Department of Ener-
gy, Washington, DC, USA, 1985.
bustion of jet fuels and fuel oils, 23 the net reduction 1ZA.F. Turhollow, T.R. Curlee and Sujit Das, Documentation of
in particulate emissions may be less than anticipated the Petroleum Allocation (PAL) Model, Oak Ridge National
by environmental planners. The final recommenda- Laboratory, Oak Ridge, TN, USA, 5 March 1987.
13Refinery Evaluation Modeling System (REMS) Documentation,
tion is that the proposals for ULS diesel fuel be DOE/EIA-0460, Energy Information Administration, October
assessed for the net reduction in particulate emis- 1984; Refinery Evaluation Modeling System (REMS) Documenta-
sions, accounting for the potential transfer of the tion, DOE/EIA-0461, Energy Information Administration, Octo-
ber 1984.
particulate emissions from highways to airways and a4REMS 1982 Verification Test Results, working paper prepared
to communities which use No 2 fuel oil for residen- for US Energy Information Administration and the Strategic
tial and commercial heating. Petroleum Reserve Office of the US Department of Energy by
Decision Analysis Corporation with Sobotka and Company,
under subcontract with Computer Data Systems, November 1983.
Research sponsored by the Mobility Fuels Technology Program 15G.T. Kinney, 'EPA to limit summertime gasoline RVP', Oil &
of the US Navy Energy and Natural Resources Office, Office of Gas Journal, 27 March 1989, p 79.
Naval Research, Interagency Agreement No 40-1391-83 under 16S. Das et al, Navy Mobility Fuels Forecasting System Phase IV
Martin Marietta Energy Systems Inc, Contract No. DE-AC05- Report, ORNL-6517, Oak Ridge National Laboratory, Oak
84OR21400 with the US Department of Energy. Ridge, TN, USA, September 1988.
17Energy Information Administration, US Department of Ener-
1Federal Register, Vol 50, No 51, 15 March 1985, p 10606. gy, Petroleum Supply Annual 1988, DOE/EIA-0340(88)/1,
2The particulate emission standards, which apply only to heavy- Washington, DC, USA, May 1989.
duty diesel engines are: 0.6 grams per brake horsepower-hour for ]SWhen extrapolated by the distillation capacity of the entire
1988 and later model years, 0.25 grams per brake horsepower- USA relative to the study regions (ie, 1/0.66), the investment
hour (0.1 grams per brake horsepower-hour for urban buses) for estimate of the study supports the investment determined in the
1991 and later model years, and 0.1 grams per brake horsepower- petroleum refining industry survey (op cit, Ref 6). The extrapola-
hour for 1994 and later model years. Tile rule is in response to the tion, which does not account for important factors such as the
requirements of section 202(a)(3) of the Clean Air Act as total refinery process configuration, provides only a rough appro-
amended in 1977. ximation.
3Federal Register, Vol 51, No 124, 27 June 1986, p 23437. ~gW.L. Nelson, Petroleum Refining Engineering, McGraw-Hill,
4The back-end distillation temperature has also been directly NY, USA, 1958, p 139.
correlated with diesel fuel particulate emissions. 2°L.D. Palmer and B.V. Copson, 'Hydrotreatment of light cycle
50p cit, Ref 3. oil for stabilization of automotive diesel fuel', in Conference
6In a 1986 survey of 139 refineries, representing 98% of the Proceedings, 2nd International Conference on Long-Term Storage
nation's crude oil refining capacity, the average diesel fuel sulphur Stabilities of Liquid Fuels, Southwest Research Institute, San
content was 0.27 weight percent. A relatively small volume of Antonio, TX, USA, 29 July-1 August 1986, pp 902-914.
diesel fuel, which was primarily consumed in the Los Angeles 21M.W. Schrepfer, R.J. Arnold and C.A. Stansky, 'Distillate
basin area, met a 0.05 weight percent maximum sulphur specifica- stability ensured by testing, treatment', Oil & Gas Journal, 16
tion. For diesel fuels sold in accordance with the ASTM D 975 January 1984, pp 79-84.
standard, the maximum sulphur content is 0.5 weight percent. See 22D.R. Hardy, R.N. Hazlett and E.W. White, 'Assessment of
National Petroleum Refiners Association, US Refining Industry storage stability additives for naval distillate fuel', Conference
Capability to Manufacture Ultra Low Sulfur Diesel Fuels: NPRA Proceedings, 2nd International Conference on Long-Term Storage
Survey 1986, National Petroleum Refiners Association, Suite Stabilities of Liquid Fuels, Southwest Research Institute, San
1000, 1899 L Street, NW, Washington, DC 20036, USA, 1986. Antonio, TX, USA, 29 July-1 August, pp 887-901.
7The average aromatics content of diesel fuel produced by the 23See, for example, D.W. Naegeli and C.A. Moses, 'Effects of
refineries surveyed by the NPRA (op cit, Ref 6) was 34 volume fuel properties on soot formation in turbine combustion,' Society
percent. ASTM D 975 has no aromatics specification but requires of Automotive Engineers Paper No 781026, Aerospace Meeting,
a minimum cetane number of 40. Cetane number and aromatics San Diego, CA, USA, 27-30 November 1978, p 5; and V.B.
content are inversely related, and, in effect, the cetane specifica- Guthrie, Petroleum Products Handbook, McGraw-Hill, NY,
tion limits the amount of aromatics that can be blended. Unzel- USA, 1960, pp 7-29.

ENERGY POLICY June 1991 493

You might also like