WO2001048120A1

WO2001048120A1 - Fuel compositions

Info

Publication number: WO2001048120A1
Application number: PCT/EP2000/013381
Authority: WO
Inventors: Richard Hugh Clark; David Andrew Ross Jones; Trevor Stephenson; Paul Anthony Stevenson; John Francis Unsworth
Original assignee: Shell Internationale Research Maatschappij BV
Current assignee: Shell Internationale Research Maatschappij BV
Priority date: 1999-12-23
Filing date: 2000-12-21
Publication date: 2001-07-05
Anticipated expiration: 2002-06-23
Also published as: AU2844801A

Abstract

A fuel composition comprising a major proportion of a base fuel, preferably a diesel fuel, and a minor proportion of a polymeric additive, which additive is soluble in said base fuel and is present in an amount sufficient to impart to the composition a kinematic viscosity which is greater than that of said composition not comprising said additive; a method of operating a compression-ignition engine, which comprises introducing into the combustion chambers of said engine said fuel composition; a method of improving the ability of a compression-ignition engine to start when at elevated temperatures by employing said fuel composition; and the use in a compression-ignition engine of said fuel composition to improve the ability of the engine to start when at elevated temperatures.

Description

FUEL COMPOSITIONS

The present invention relates to fuel compositions, preferably diesel fuel compositions, for use in compression-ignition engines, and methods of operating said engines. As stated in O-A-95/33805 (Exxon) environmental concerns have led to a need for fuels with reduced sulphur content, especially diesel fuel and kerosene. However, the refining processes that produce fuels with low sulphur contents also result in a product of lower viscosity and a lower content of other components in the fuel that contribute to its lubricity, for example, polycyclic aromatics and polar compounds.

At present, a typical sulphur content in a diesel fuel is about 0.25% by weight (2500 ppmw) ("ppmw" is parts per million by weight) . In Europe maximum sulphur levels have been reduced to 0.05% (500 ppmw) and will be reduced to 0.035% (350 ppmw) in 2000; in Sweden grades of fuel with levels below 0.005% (50 ppmw) (Class 2) and 0.001% (10 ppmw) (Class 1) are already being introduced. As such fuels are becoming of greater interest (due to environmental concerns) the density and viscosity specifications could be taken even lower in the future.

The definition of a City Diesel fuel varies from country to country, but is most commonly defined in terms of its sulphur content (ultra low <50ppmw) and its density (<835kg/m³) . In the context of this specification City Diesel fuels are those having a sulphur level below 0.01% by weight (100 ppmw) that fall within the density range of 800 - 835 kg/m³. Low viscosity City Diesel fuels can have potential problems wherein vehicles may become susceptible to hot starting problems, especially during hot weather. The vehicle may stall at idle and fail to restart unless the fuel pump is cooled. It has now been surprisingly found that use of polymeric additives, which may comprise one or more polymers, in diesel fuels can increase the viscosity of diesel fuel compositions and dramatically reduce the occurrence of hot starting problems in vehicles using said compositions, whilst leaving other fuel specification properties (e.g. sulphur and density) unaffected.

According to the present invention there is provided a fuel composition for a compression- ignition engine comprising a major proportion of a base fuel, preferably a diesel fuel, and a minor proportion of a polymeric additive, which additive is soluble in said base fuel and is present in an amount sufficient to impart to the composition a kinematic viscosity which is greater than that of said composition not comprising said additive. In another aspect, the present invention provides a method of operating a compression-ignition engine, which comprises introducing into the combustion chambers of said engine a fuel composition according to the invention as defined above. The present invention further relates to the use in a compression-ignition engine of said fuel composition to improve the ability of the engine to start when at elevated temperatures, as compared to such ability when the engine is operated using said composition not comprising said polymeric additive.

The present invention also further provides a method of improving the ability of a compression-ignition engine to start when at elevated temperatures by employing a fuel composition according to any one of Claims 1 to 7, as compared to such ability when the engine is operated using said composition not comprising said polymeric additive.

City Diesel fuels tend to have lower sulphur content and density than conventional diesels in order to produce lower vehicle emissions. The property of fuel density is strongly linked with that of fuel viscosity, thus fuels of low density tend to have low viscosity. Low viscosity fuels can have potential problems wherein vehicles have shown themselves to be susceptible to hot starting problems, especially when using these City Diesel fuels during hot weather.

Middle distillate fuel oil can be derived from petroleum and typically has a boiling range in the range 100°C to 500°C, e.g. 150°C to 400°C. Such petroleum- derived fuel oils may comprise atmospheric distillate or vacuum distillate, or cracked gas oil or a blend in any proportion of straight run and thermally and/or catalytically cracked distillates: Preferred fuel oil compositions of the invention are diesel fuel compositions . Middle distillate fuels typically have an initial distillation temperature of approximately 140°C to 210°C and final distillation temperature of approximately 250°C to 390°C, depending on fuel grade and use. Diesel fuels typically used in road-going vehicles propelled by compression-ignition engines have an initial distillation temperature of approximately 150°C to 210°C and final distillation temperature of approximately 280°C to 390°C. The diesel fuel itself may be an additised

(additive-containing) fuel or an unadditised (additive- free) fuel. If the diesel fuel is an additised fuel, it will contain minor amounts of one or more additives, e.g. one or more additives selected from anti-static agents, pipeline drag reducers, flow improvers (e.g. ethylene/vinyl acetate copolymers or acrylate/maleic anhydride copolymers) and wax anti-settling agents (e.g. those commercially available under the Trade Marks "PARAFLO " (ex Infineum International Ltd.), "DODIWAX" (ex Clariant GmbH) .

The diesel fuel has a sulphur content of at most 0.2% by weight (2000 ppmw). Advantageous compositions of the invention are also attained when the sulphur content of the diesel fuel is below 0.05% by weight (500 ppmw), preferably below 0.005% by weight (50 ppmw) or more preferably below 0.001% by weight (10 ppmw) . The diesel fuels should fall within the density range of from 800 - 845 kg/m³, preferably in the range of from 800 - 835 kg/m³, more preferably in the range of from 800 - 820 kg/m³. The viscosity of the diesel fuel should be in the range of from 1.3 - 2.7 cSt @ 40 °C, preferably in the range of from 1.5 - 2.5 cSt @ 40 °C. Other desired properties for said diesel fuels are disclosed in the European CEN 590 specification.

The definition of the fuel -additive blend viscosity is not straightforward, in that the additives shear when passed through mechanical equipment (e.g. fuel injection equipment, FIE) . Thus one can define both a "sheared" and an "unsheared" viscosity for a given fuel/additive combination. The shear that the fuel/additive blend is subjected to in an engine fuel system will decrease its viscosity. In practice, the "sheared" viscosity is a more sensible number to use when rating fuel performance, because the fuel will undergo a degree of re-circulation and shearing within the fuel system. Thus, sheared viscosity is more representative of the viscosity of the fuel in a real fuel system.

Diesel fuel compositions according to the present invention have a kinematic viscosity ("sheared viscosity") of at least 2.0 cSt at 40°C, measured following shearing caused by use in a compression- ignition engine. The sheared viscosity of said compositions will preferably be at least 2.5 cSt at 40 °C, more preferably at least 2.7 cSt at 40 °C. The sheared viscosity may be determined as outlined in the Examples below.

Polymeric additives which may be employed in diesel fuel compositions according to the present invention have a weight average molecular weight, M„, preferably in the range of from 0.1 million to 20 million, more preferably in the range of from 0.3 million to 20 million, more preferably in the range of from 0.5 million to 20 million, more preferably in the range of from 0.5 million to 8 million, and even more preferably in the range of from 0.5 million to 5 million, and/or a viscosity average molecular weight, Mvis, preferably in the range of from 0.1 million to 20 million, more preferably in the range of from 0.3 million to 20 million, more preferably in the range of from 0.5 million to 20 million, more preferably in the range of from 0.5 million to 8 million, and even more preferably in the range of from 0.5 million to 5 million.

The weight average molecular weight, M , of the polymeric additive may be determined by several techniques which give closely similar results. Mw may be easily determined by gel permeation chromatography (GPC) with calibration of the polymer, as will be appreciated by those skilled in the art. Alternatively, M_w may be determined by light scattering, as described in ASTM D 4001-93.

The viscosity average molecular weight, Mvis, of the polymeric additive may be conveniently determined by the same viscosity experiments used to determine intrinsic viscosity [η] (see below) . ^,. of a polymer is related to intrinsic viscosity via the equation: -

[η]=KM_vis ^a The constants "K" and "a" depend on the solvent and solvent type and are determined from experiments on standard nearly monodisperse solutions.

Polymeric additives which may be employed in diesel fuel compositions according to the present invention are those which whilst being soluble and inert in the fuel composition also increase the viscosity of said composition. Suitable polymeric additives include, but are not limited to, polyalkenes and polyethers . Suitable polyalkenes include polymers of monoolefins, diolefins and cyclic olefins. The polyalkenes may conveniently be homopolymers or copolymers, for example of at least one C2-C10 monoolefin, such as ethylene, propene, 1-butene, 1- hexene and 1-octene. Preferably the polyalkene is a polymer of at least one C2-C5 monoolefin, e.g. an ethylene-propylene copolymer. The monoolefin is preferably a C3-C4 olefin and preferred polyalkenes derived therefrom include polyisobutylenes and atactic or isotactic propylene oligomers. The polyalkene may be a polymer of conjugated diene monomers such as isoprene or butadiene or it may be a copolymer of conjugated dienes with other co- polymerisable monomers, for example a mono vinyl aromatic hydrocarbon. Examples of the mono vinyl aromatic monomers may be selected from styrene, α-methylstyrene, p-methylstyrene, m-methylstyrene, o-methylstyrene, p-tert-butylstyrene, dimethylstyrene, and various other alkyl-substituted styrenes, alkoxy-substituted styrenes, vinylnaphthalene, and vinyl xylene . The alkyl and alkoxy groups of the alkyl -substituted or alkoxy substituted styrenes respectively may comprise from 1 to 6 carbons, preferably from 1 to 4 carbons. Of these monomers styrene is the preferred vinyl aromatic monomer. The conjugated diene monomers are conveniently dienes with from 4 to 8 carbon atoms per monomer and may be selected from butadiene, isoprene, 2 -ethyl -1,3- butadiene, 2 , 3-dimethyl-l, 3 -butadiene, 1, 3 -butadiene, 1, 3-pentadiene, 2 , 4-hexadiene, 3 -ethyl- 1, 3-pentadiene, and mixtures thereof .

Of these monomers styrene and butadiene or isoprene, or mixtures thereof are preferred. Most preferable are block copolymers which contain only substantially pure poly (butadiene) or pure poly (isoprene) blocks.

Polyalkene copolymers according to the invention may be random and/or block copolymers . The block copolymers may be linear triblock or multiblock copolymers or multi -armed or star shaped symmetrical or unsymmetrical block copolymers.

Poly (conjugated diene) blocks may be completely, partially or selectively hydrogenated.

The olefin monomer is most preferably isobutylene, isoprene or butadiene, so that polyisobutylenes, polyisoprenes and polybutadienes are the preferred forms of polyalkene .

Examples of suitable commercial polyisobutylenes are those sold by BASF under the trade designations "Oppanol B100" and "Oppanol B200" . Examples of suitable commercial polyisoprenes are those sold by member companies of the Royal Dutch/Shell Group under the trade designations of "Cariflex IR-307" and "Cariflex IR-500" .

Examples of suitable commercial polybutadienes include that sold by Dow Chemical under the trade designation "BR1202G" . Suitable polyethers according to the invention include aromatic polyethers, and polymers of ethylene oxide, propylene oxide and higher 1 , 2-epoxides . Examples of suitable commercial polyethers include polyvinyl isobutyl ether polymer available from Aldrich (catalogue number 20,034-4).

Suitable polyalkenes and polyethers according to the invention may additionally contain polar substituents . The amount of polymeric additive incorporated in the diesel fuel composition is in the range of from 200 to 10000 ppmw based on said fuel composition. Preferably, the polymeric additive will be present in an amount in the range of from 250 to 5000 ppmw, and more preferably in the range of from 500 to 5000 ppmw. In the present specification, by "elevated temperature" is meant a fuel return line temperature of at least 45°C.

The present invention is illustrated by the following Examples, which should not be regarded as limiting the scope of the invention in any way. In the following description, all parts and percentages are by weight, unless stated otherwise, and temperatures are in degrees Celsius.

The Limiting Viscosity Number (LVN) or intrinsic viscosity [η] is defined for a given polymer type in a given solvent at a defined temperature. For a set of similar polymer types, their LVNs in the same solvent are an indicator of the respective molecular weights of the polymers . The intrinsic viscosity may be conveniently determined by viscosity experiments as described in Physical Chemistry P. . Atkins, 5^th Edition, Oxford

University Press, 1993, pp 784-786 and 798-799. From experimental viscosity data of polymer solutions at different concentrations and use of the equation defining intrinsic viscosity [η]:-

[η] =ϋm_c→0 (η/η^*-l) , c where η is the viscosity of a polymer solution of concentration c, and η* is the viscosity of the pure solvent , both measured at the chosen temperature . The value η/η* is referred to as the "viscosity ratio" .

Plots of either (η/η*-l)/c or (In (η/η*) ) /c versus c, can be extrapolated back to c=0 to derive a value for the intrinsic viscosity [η] . An article describing the development of a procedure for the determination of intrinsic viscosity from the viscosity ratio measurement at only one concentration is described in the reference: S.H. Maron, J. Appl . Pol.

Sci. 5 (1961) 282. In the examples, the following polymeric additives were used:

Polymer A is a solution polymerised polyisoprene polymer with a high cis-1,4 content and a high molecular weight as indicated by its nominal limiting viscosity number of 7.75 dl/g. M„ is in the range 0.3 million to 5 million. The material is available under the trade designation "Cariflex IR-307" from member companies of the Royal Dutch/Shell Group.

Polymer B is a solution polymerised polyisoprene polymer with a high cis-1,4 content and a high molecular weight as indicated by its nominal limiting viscosity number of 9.25 dl/g. , is in the range 0.3 million to 5 million. The material is available under the trade designation "Cariflex IR-500" from member companies of the Royal Dutch/Shell Group. Polymer C is a solution polymerised hydrogenated isoprene-styrene "star" polymer and is available under the trade designation "Shellvis 260" from Infineum

International Ltd. M_w = 0.9 million.

Polymer D is a solution polymerised polyisobutylene polymer (N^ = 1.11 million) available under the trade designation "Oppanol B100" from BASF.

Polymer E is a solution polymerised polyisobutylene polymer (M^ = 4 million) available under the trade designation "Oppanol B200" from BASF. Polymer F is a solution polymerised polyvinyl isobutyl ether polymer available from Aldrich (catalogue number 20,034-4). M,, = 0.6 million.

Polymer G is a solution polymerised polybutadiene polymer available under the trade designation "BR1202G" from Dow Chemicals. M„ = 0.38 million.

Note : M„= weight average molecular weight M^ = viscosity average molecular weight.

EXAMPLES

Example 1 The example describes tests of the above materials dissolved in automotive gas oil (diesel) fuels to examine their effectiveness in overcoming hot starting problems in a bus using "City Diesel".

The test which follows was designed to measure the effects of fuels and additives on hot starting in a bus by reproducing, in a repeatable manner, the conditions that exist in the engine which cause the hot starting problems .

The test used a Dennis Dart single decker bus fitted with a Cummins 6BT5.9 engine and a Lucas CAV DPA distributor pump. The characteristics of the bus are shown in Table 1.

Table 1 : Description of Dennis Dart bus

The bus was run on a chassis dynamometer in a climate controlled chamber at an ambient temperature of 30+l°C (representative of the maximum ambient summer temperature) . The bus was installed on the dynamometer and a road load model equivalent to rolling resistance plus an inertia of half maximum payload was defined. These data were taken from typical values used during emissions testing of buses. A large fan was positioned in front of the bus and the air flow directed under the bus to simulate the normal air flow when driving. A smaller fan was positioned at the rear offside to provide some air flow around the radiator vent . The engine compartment door was kept closed during testing. The vehicle's own fuel tank has a capacity of 120 litres. Use of a small fuel can was preferred in order to maximise the rise in fuel temperature due to fuel returning to the tank. Thus, 25 litre fuel cans were connected to feed and return lines. The fuel system was purged with the test fuel by feeding fuel from one can and allowing the returned fuel to run to waste in another can. When enough fuel had been bled through, the return line was switched to the supply can. Engine speed was measured from the bottom pulley, using a number of reflective strips and a detector. Temperatures were measured using K-type thermocouples at the following locations: fuel tank (or can) ; Inlet manifold air temperature (via blanking plug in the manifold) ; temperature of fuel returned from engine (at the bottom of the engine compartment) ; engine compartment air temperature (near the fuel pump) ; coolant temperature (top hose at exit from the engine block) ; sump oil temperature (via dipstick) ; ambient air temperature . Two base fuels were used in the tests both of which had viscosities near or below the current European CEN EN 590 limit of 2.0 cSt . Fuel A was a Swedish Class 1 diesel fuel, Fuel B was an odourless kerosene. The fuel properties are listed in Table 2. Table 2 : Properties of Fuels

Fuel C, which had a moderate viscosity without the use of additives was used to benchmark the characteristics of the bus engine during the hot start trial. With this fuel, the engine completed five of the five minute runs at 30 mph, before failing to start. The data has been plotted on Figure 1, showing the decrease in viscosity of the fuel as the fuel return line temperature increases. The pass/fail criterion is achieved when the fuel return line temperature is between 51 - 54 °C or the viscosity crosses the 2.15 cSt border at this temperature.

From the results of this unadditivated fuel, one can define the pass/fail criterion for a fuel. Fuel C can be defined as having adequate hot start performance in this particular bus engine. Thus any fuel which achieves its viscosity (Vk40) by (i) inherent viscosity, (ii) unsheared additive or (iii) sheared additive, and that viscosity equals or exceeds 2.7 cSt (at 40°C) , can be defined as a fuel of acceptable hot start performance for this example. Fuel/additive blends of 25 litres were prepared by weighing out the desired amount of polymer and then dissolving it in batches in the stirred fuel . Polymers A and B are rubber-like materials and were cut into small pieces (approximately O.lg) prior to dissolution. Polymer C is a powder. The fuel blends prepared are listed in Table 3. The viscosity of some of the fuel blends were not measured directly but were calculated assuming a linear relationship between viscosity and additive concentration. Some fuels have two or more viscosities .quoted against them, this reflects the fact that the fuel/additive combinations will shear within the components of an engine and thus the viscosity will decrease, this is discussed further in Example 2.

Table 3: Fuel/additive blends, pass/fail criterion for minimal engine running (low shear) i

Vk40 > 2.7cSt

0\

IP71/ASTM D445, ND = not determined

A low fuel volume was used during testing to simulate the situation towards the end of a working day when the fuel tank is nearly empty and the remaining fuel is heated by the engine return. Tests started with 15 litres. Bus and tests fuels were left to soak overnight in the dynamometer at an air temperature of 30°C. At the end of the day the fuel system was prepared with the first fuel to be tested on the next day if convenient. The first test was able to start with a fuel temperature of 30°C but subsequent tests always began with some residual heat from the previous run.

The following procedure was used to assess hot restarting and idle speed: 1. The fuel system was purged with the new fuel, allowing 5 litres to run to the waste container before redirecting the return fuel line to the supply container. The engine was allowed to cool for as long as possible before the first engine start attempt .

2. The recording of vehicle speed and all temperatures at 1 Hz frequency was started.

3. The engine was started by cranking for up to 10 seconds. If the engine did not fire on the first attempt there was a 10 second wait and then it was cranked again for up to 10 seconds. The total cranking time was recorded using a stopwatch (+ 0.1s), and the number of attempts also recorded.

4. When the engine started it was allowed to idle for 60 seconds and the idle speed recorded. 5. The bus was then accelerated to 30 mph and any driveability problems noted, e.g. stall, lack of power, roughness .

6. The bus was held at a constant 30 mph for 5 minutes. 7. The bus was brought to rest by braking and the engine allowed to idle for 60 seconds. The idle speed was recorded.

8. The engine was switched off.

9. After a 10 second wait steps 3 to 8 were repeated until the engine failed to start or until the return fuel temperature stopped rising.

10. The recording of temperatures and vehicle speed was stopped.

11. A 150 ml sample of fuel was collected from the return line and a 75 ml sample from one of the fuel injector lines.

12. The bus was stopped and changed to the next test fuel . The engine compartment was cooled with a compressed air jet. Notes : i. If the engine failed to restart in step 3 after 6 cranking attempts the engine was allowed to cool with the compartment door open for 5 minutes . 3 further attempts were made to start the engine. If the engine still did not start, the test proceeded to the next fuel . ii. The battery of the bus was re-charged overnight to avoid it being run down.

The first fuel tested was Fuel A (Vk40 = 2.01 cSt) which was put into the bus before the overnight soak. The first starting attempt was therefore at a fuel temperature of 30°C. The engine did not start or idle with this fuel .

The next fuel tested was Fuel Blend J, a blend of Fuel A with 1100 ppmw Polymer C (Vk40 = 2.295 cSt) . This fuel had a higher viscosity than Fuel A but the engine still did not start or idle properly.

The next fuel tested was Fuel Blend K, a blend of Fuel A with 2200 ppmw Polymer C (Vk40 = 2.642 cSt) . The engine started with difficulty (using an ether spray into the air intake) but this could have been because of traces of the previous fuel remaining in the fuel pump. The engine was run at 30 mph and it was noted that the engine ran better at idle. The first starting attempt was made when the return fuel line temperature reached 35°C. The engine failed to re-start.

The engine was allowed to cool as much as possible before testing with Fuel Blend F (Fuel A with 3200 ppmw Polymer A) (Vk40 = 5.055 cSt) . The engine started successfully and a total of six 5-minute runs at 30 mph were made before the engine failed to start .

Fuel Blend H was tested next (Fuel B with 4600 ppmw Polymer B) (Vk40 = 4.078 cSt) . Again the engine was cooled as much as possible before the first starting attempt. The engine started with difficulty and idled roughly. Slight pressure on the accelerator pedal was sufficient to restore a smooth idle. It seems likely that the low density of this fuel was causing slight under-fuelling at idle. The engine completed three 5- minute runs at 30 mph before failing to start. Fuel Blend E (Fuel A with 1550 ppm Polymer A) (Vk40 = 3.322 cSt) was tested and completed four 5-minute runs at 30 mph before failing to start.

Thus in Example 1, the concept of overcoming hot start problems by using fuels with viscosity additives was proven. It is evident that there are a range of additives that are appropriate, provided that they can increase the viscosity of the fuel sufficiently. Example 2 When assessing the viability of fuel/additive combinations to overcome hot start problems, a convenient, repeatable and universal measure of "sheared" viscosity is required. To that end, the CEC mechanical shear stability test (CECL-14-A-88) is a recognised industrial standard for the shearing of lubricants and it is particularly appropriate to apply to diesel fuels because it is based on a diesel fuel injection system. The test fluid is pumped from a reservoir though a diesel injector and then re- circulated back to the reservoir. For lubricants, the standard test condition is 60 cycles (re-circulations) . However for fuel/viscosity additive combinations, an appropriate test condition had to be found.

A chosen fuel/additive combination was split into a variety of batches, each batch was subjected to a different number of cycles in the CEC test rig, i.e. 2, 4, 8, 16, 32 cycles. The sheared viscosity for each of these conditions was recorded. The results from this CEC shearing rig were compared with a more realistic full scale fuel injection pump rig based on a Volkswagen vehicle. The CEC test run on 4 cycles gave the best agreement with the Volkswagen rig (Figure 2) , on which experiments had shown that 50% loss of viscosity was a suitable level.

A large number of additive types/structures were then tested using the CEC rig at the chosen test conditions. These blends had both their sheared and unsheared viscosity determined. From the results it was evident that the sheared viscosity always exceeds the base fuel viscosity, so that all these additives would be appropriate to use in the prevention of hot start problems, although the actual concentration levels would have to be tailored.

In this shear rig test, the test fuels were identical to those in Example 1 (see Table 3) . Two base fuels were used in the tests both of which had viscosities near or below the current CEN EN 590 limit of 2.0 cSt . Fuel A was a Swedish Class 1 diesel fuel and Fuel B was an odourless kerosene.

The polymeric additives used in these tests have been described previously (polymeric additives A, B and C were identical to those described in Examples 1) .

If a fuel is blended with an unsheared additive to give a Vk40 just above 2.7 cSt, then as time passes and re-circulation occurs, the additive will shear and the fuel blend's Vk40 will fall below 2.7 cSt . Thus for any practical application of these additives, the fuel blend Vk40 should still exceed 2.7cSt even after the fuel blend has endured a typical level of shear within an engine . From the above, it is apparent that the sheared viscosity of a fuel blend is a better criterion of its hot start performance than its unsheared viscosity, when considering the practical application of these additives in the field. To that end, in assessing whether a particular fuel blend/additive combination is suitable to prevent hot start problems, a quick and simple laboratory-based method is needed to determine sheared viscosity. There will not be a one-to-one correspondence between the field trials of Example 1 and the mini shear rig results of Example 2. This is because the field trial results are done under relatively mild conditions of shear. That is, the field trial is performed immediately after the introduction of the fuel, so that a minimal quantity of the fuel sample would have been re-circulated and sheared. Thus the viscosity will only have declined minimally from the unsheared viscosity. In contrast, the mini shear rig conditions were chosen to represent severe conditions where the majority of the fuel has been re-circulated and the tank is nearly empty, resulting in the high quantity of fuel being sheared. This severe condition defines a "fail-safe" criterion.

Thus, the aim of the mini shear rig was to provide a measurement of both sheared and unsheared viscosity (Vk40) for candidate viscosity additive fuel blends, the sheared value providing a fail-safe level of viscosity as described above .

In this given bus, sample fuel blends will work initially, provided their Vk40 (of unsheared fuel) > 2 . 1 cSt at the start of testing. However, after several hours of bus operation and fuel re-circulation, the necessary criterion is for the sheared viscosity

> 2.7cSt.

Using the marker Fuel C, which defines a Vk40>2.7cSt as a fuel which will pass the hot start trial, then the data from the mini rig in Table 4 can be re-worked to see whether this large number of additive/fuel combinations would have given a successful outcome if used in the bus field trial.

Condition TAl At the start of bus running, the fuel blends would have undergone little shear, so provided the Vk40 of unsheared fuel is > 2.7cSt, a pass would have been achieved in the bus hot start test .

Condition fBl However, under the most severe conditions shear expected, the operative figure is the sheared viscosity being > 2.7 cSt, here 8 of the 16 blends would fail in this particular bus.

Table 4 Polymer/Fuel blends used in the CEC mechanical shear stability test (CECL-14-A-88) (all subjected to 4 cycles)

Claims

C L A I M S

1. A fuel composition for a compression-ignition engine comprising a major proportion of a base fuel, preferably a diesel fuel, and a minor proportion of a polymeric additive, which additive is soluble in said base fuel and is present in an amount sufficient to impart to the composition a kinematic viscosity which is greater than that of said composition not comprising said additive.

2. A fuel composition according to Claim 1, having a kinematic viscosity of at least 2.0 cSt at 40°C, measured following shearing caused by use in a compression-ignition engine.

3. A fuel composition according to Claim 1 or 2 , wherein said additive has a weight average molecular weight in the range of from 0.1 million to 20 million.

4. A fuel composition according to Claim 1 or 2 , wherein said additive has a viscosity average molecular weight in the range of from 0.1 million to 20 million.

5. A fuel composition according to any one of the preceding claims, wherein said additive is selected from polyalkenes and polyethers .

6. A fuel composition according to any one of the preceding claims, wherein said additive contains polar substituents .

7. A fuel composition according to any one of the preceding claims, wherein said additive is present in an amount of from 200 to 10000 ppmw based on said fuel composition.

8. A method of operating a compression-ignition engine, which comprises introducing into the combustion chambers of said engine, a fuel composition according to any one of Claims 1 to 7.

9. A method of improving the ability of a compression- ignition engine to start when at elevated temperatures by employing a fuel composition according to any one of Claims 1 to 7, as compared to such ability when the engine is operated using said composition not comprising said polymeric additive.

10. The use in a compression-ignition engine of a fuel composition according to any one of Claims 1 to 7, to improve the ability of the engine to start when at elevated temperatures, as compared to such ability when the engine is operated using said composition not comprising said polymeric additive.