DIESEL COLD FLOW PROBLEMS AND EARLY WARNING INDICATORS
Posted: August 1, 2014
by Stu Porter
Quality Control Co-ordinator for Suncor Energy Products Inc.
It is a well-documented fact that diesel fuel and engine can experience weather-related difficulties during winter months in cold geographical areas. The literature and information readily found in the public domain (for example, on the World Wide Web) are replete with incidences whereby diesel operators have trouble starting their engines during cold mornings, or have found their fuels sufficiently gelled and fuel filters plugged that their vehicles are rendered non-operational. Not surprisingly, there exists an equally diverse amount of practical advices and commercial products meant to minimize or even eliminate these problems.
This article focuses on a major component of the cold-weather difficulty: that of the diesel fuel itself. The cold flow behavior of diesel fuel will be illustrated, and standardized tests to characterize its cold flow property will be described and compared. The value of these tests in providing a practical early warning to diesel fuel users of an impending gelling problem will be explained. Finally, the various solutions proposed by fuel suppliers, users, and after-market third-party vendors will be discussed.
Diesel Fuel and low temperature operability
All diesel fuels (#2 grade diesel in the present context) consist of a multitude of hydrocarbon compounds of differing molecular weights [see, e.g. ref (1), chap.4]. When a fuel is being cooled, as in cold ambient temperatures during winter time, the heavier components will typically solidify, in much the same way as water freezes into ice. This solid component, commonly called paraffin waxes, will continue to grow and aggregate into larger pieces (gelling) given sufficient time or if the ambient temperature continues to fall. These waxes can precipitate out of the liquid fuel and adhere onto cold surfaces in the surroundings. This is particularly harmful for fuel filter as the waxes can block the small openings, thereby plugging the filter. Even if the waxes were to remain suspended in the liquid fuel, if sufficiently large numbers of such exist due to prolonged exposure to cold weather, their sheer number can still effectively restrict free flow of fuel, hence starving and stalling the engine.
Fuel Specification and Testing
That diesel fuel can suffer from gelling problems, affecting its utility and operability at cold temperatures, is well-known to fuel producers and suppliers. Accordingly, there exist a number of test procedures to assure fuel quality and aid information exchange. Such metrics as the cloud point, pour point, and filter operability tests have been used by petroleum refineries and adopted by international standard bodies to denote various aspects of diesel fuel when subjected to cold temperatures. Their meanings and usages will be explained next.
The cloud point (CP) of a fuel is defined by ASTM (American Society for Testing and Materials) as the temperature at which waxes appear upon cooling under prescribed conditions of the test method. The sample is descriptively said to form a “cloud”. In short, above its CP, a fuel will not form waxes that lead to gelling/plugging difficulty. Below its CP, components within a fuel will begin to solidify and can potentially lead to engine problem. From this viewpoint, CP is an absolute “fail-safe” limit.
The pour point (PP) is the lowest temperature at which a fuel will move (“pour”) under prescribed conditions of the test method. The mechanics causing the fuel not to move (pour) is due again to the continual formation and subsequent binding of waxes into a macroscopic gel, thereby trapping and preventing the remaining liquid fuel from moving. The utility of the PP coincides with the fuel’s pumpability and handling characteristic; it is less useful when intervening meshes and filters are present and hence will not be discussed further.
Operability metrics such as the cold filter plugging point (CFPP) or low temperature filterability test (LTFT) physically introduce wire meshes in the test methods. The CFPP or LTFT are the lowest temperature at which a fixed amount of fuel successfully passes through the mesh within a pre-determined amount of time, again under respective prescribed conditions of the methods. In some quarters, these tests represent more “realistic” simulations of the interaction between fuel and engine components with the imposition of the wire meshes.
The provisional clause “under prescribed conditions of the test method” is revealing and deserves further discussions. Whether with the CP, CFPP, or LTFT test, in each case the sample is under a dynamic condition, i.e. in the process of being cooled, when the metric is determined. As a result, the CP (or CFPP or LTFT) is meaningful, contingent upon the cooling rate and other conditions (e.g. mesh size, pump pressure) of the test. What the result purports to represent, and how well it relates to actual field conditions, should then be interpreted and judged accordingly.
To take a concrete example: CP is the temperature when a cloud in the sample is first observed as the sample is being cooled at specified rates. It does not, however, indicate the amount and extent of waxing were the sample kept at the same temperature for an extended period of time, thus allowing the wax crystals to grow and agglomerate. Consequently, there is a general perception that although waxes exist at the CP, the amount is typically insignificant to affect fuel operability and engine performance. This impression can be misleading as shown below.
Figure 1 shows an ASTM reference diesel fuel (DL0002) at -13°C, or 1°C above its inter-laboratory consensus CP of -14°C. No wax is observed and the figure displays a plain dark background. Figure 2 shows the same fuel at 15 seconds after reaching the CP (of -14°C) and being held at this temperature. One can detect the presence of a few crystals, identified as bright “spots” in the figure. The appearance of waxes, albeit in small amount, coincides with the definition of CP as expected. By contrast, the extent of wax formation in the same sample after being held for 60 minutes at the CP is shown in Figure 3. Here, an extensive amount of clouding permeates the whole sample, with crystal sizes also larger than those in Figure 2. The kinetic effect of long time exposure, even at a temperature as seemingly innocuous as the CP, is clearly displayed here. The likelihood of filter plugging and engine failure is correspondingly higher.
A similar experiment involving a diesel treated with a cold-flow additive is equally instructive. The sample originates from a CEN (European Committee for Standardization, or Comité Européen de Normalisation) 1999 CFPP inter-laboratory study [ref (2)] and has an average CFPP value of -3°C. Figure 4 indicates the appearance of the fuel after being held at the CFPP for 15 seconds. While there is a general haziness or cloudiness, individual crystals are small in sizes. However, after the sample has been held at the CFPP for a longer period of 60 minutes, the amount of crystal growth is much more extensive, as shown in Figure 5. Again the effect of time exposure may alter the wax habit and increase the probability of filter plugging. It should be noted that CFPP is checked while the sample temperature is dropping; not after a specified holding period. For this reason many have pegged the difference between CP and CFPP of their diesel specification to be within a certain limit. For example, one refiner has specified their diesel’s CFPP to be no greater than 8°C below the CP, in order to “ensure there is not an excess of fine wax crystals which may be sufficient to block filters” [ref (2)].
Therefore one needs to exercise acumen when scrutinizing the results of filter operability tests. Here the relevant considerations are cooling rates as well as mesh sizes. CFPP employs a generally rapid rate of about 40°C per hour [ref (1), chap. 5] and a wire mesh size of 45 microns (µm). In contrast, LTFT cools at a much slower 1°C per hour with a mesh size at 17µm. Consequently, CFPP is a relatively quick test (typically less than one hour), while LTFT typically requires 24 to 48 hours to complete, making it impractical to use in the field. On the other hand, it can be justifiably argued that the slower cooling rate is a more realistic emulation of conditions in the field, allowing for the effects of prolonged soaking, and that the smaller mesh size is more representative of the fuel filter sizes commonly used in diesel vehicles. It may be remarked that newer generation diesel vehicles typically employ primary (at the fuel pump) filter sizes between 10 to 20 µm, and secondary (at the engine) filter sizes of 2 to 5 µm [ref (3)].
In a pioneering 1981 CRC (Coordinating Research Council) study [ref (4)] on the use of CP, CFPP, and LTFT versus the drivability of trucks, it is found that CP is the most robust and conservative measure at 98% in correctly predicting the minimum operating temperature of the test vehicles, while LTFT provides correct prediction 89% of the time, and CFPP 64% of the time. It should be noted that test with prediction capability near 50% means the test is virtually useless as flipping of a coin will in the long run achieve a 50% probability of success.
A more recent non-independent study conducted in 2000 [ref (5)] using newer-generation heavy-duty vehicles duplicates many of the previous results: at low ambient temperatures, LTFT protects the test vehicles 78% of the time, while CFPP protects the test vehicles 65% of the time. The report further expresses concerns that the trend towards smaller porosity filters in some diesel trucks render them more prone to filter plugging problems.
Although all of the above test methods are recognized by international standard organizations and accorded with test method designations [ref (6)], some are specifically allowed in specifications used between buyers and sellers, or recommended in consensus positions by other regulatory agencies or trade interest groups. For example, ASTM D975 (“Standard Specification for Diesel Fuel Oils”) [ref (7)] provides a guideline in low temperature operability requirement, allowing only CP be used as the test parameter to conform to the anticipated ambient weather temperatures encountered by the users. On the other hand, both CP and LTFT are endorsed by NCWM (National Conference of Weights and Measures) in its specification for premium diesel [see, e.g. ref (1), chap. 5].
The equivalent European Union specification, CEN EN-590 [ref (8)], specifies both CP and CFPP in its winter grade (arctic) diesels, but limits CFPP to be no more than 10°C below the CP. This is similar to the 2000 World-Wide Fuel Charter [ref (9)], a set of harmonized recommendations for fuel quality under the auspices of ACEA (European Automobile Manufacturers Association), the Washington-based Alliance of Automobile Manufacturers, the Chicago-based EMA (Engine Manufacturers Association), and finally JAMA (Japanese Automobile Manufacturers Association). The Charter therefore represents the consensus of most of the vehicle and engine manufacturers and trade associations in the world. It allows all of CP, LTFT, and CFPP to serve as compliance parameter, but again limits the CFPP to be no lower than CP by 10°C. It rationalizes this point by stating that “CFPP alone will not fully describe the cold flow performance. CP is needed in support, for adequate correlation between measured CFPP value and real vehicle operability” [ref (9)].
A widespread filter-plugging problem last winter in New Zealand involving over 16,000 financial claims [ref (10)] has prompted the government to change regulations on cold-flow tests. In their recently approved 2002 regulations [ref(11)], winter diesels must now be tested by both CP and CFPP instead of the earlier regulations which allowed either CFPP or CP be used to release products. Also, the new regulations limit CFPP to be no lower than CP by 8°C.
Other organizations have actually preferred some test methods over others. In the joint 2002 EMA/TMC (The Maintenance Council) Consensus paper [ref (12)], it is stated that “LTFT provides the best overall correlation with field performance. However, for non-additized fuel, CP and LTFT correlate very well. Since CP is more practical as a quality control test, it is listed as the primary recommendation.” Similar recommendations recur in another of EMA’s Consensus Position on premium diesel fuel (FQP-1A) [ref (13)].
Methods to address Low Temperature Operability available to the Suppliers
Several avenues are available to the fuel producers and suppliers to ensure acceptable low temperature performance in their diesel fuels. Although not all refiners can take advantage of these methods equally, considerable flexibility can be found in the choice of the underlying crude oil, the refining method, blending with lighter (lower CP and less waxy) components, or treating with appropriate additives, without violating other specifications. In addition, some or all of the above-mentioned low-temperature test methodologies (CP, CFPP, LTFT) would be used to monitor the fuels during production and to certify them during final release and custody transfer.
Naturally, the attainment of low temperature performance specifications cannot be achieved without incurring compromises in other areas, whether in cost, fractional yield of other types of petroleum products, or other properties characteristic to the diesel fuel itself. For example, blending with lighter components typically lower the energy value, which often translates to worse fuel economy. Additization has been used primarily to modify the shape, size, and degree of aggregation of wax crystals, which is helpful in lowering the CFPP and LTFT compared to an un-additized fuel, but seldom can it alter the CP. In some cases, improving CFPP or LTFT means the low temperature operability of diesel vehicles is enhanced; but in other cases, the relationship does not exist due to variations in vehicle design. Furthermore, it has been cautioned that the “interactions (between fuel and additives) are fairly specific, so a particular additive generally will not perform equally well in all fuels” [ref (1), chap.7]. In any case, that the low temperature performance characteristic is fully verified before the product is released, must be a fundamental expectation by the users towards the suppliers.
Factors of Uncertainty
However, there remain three considerations beyond the capabilities of the refineries.
The first is the more pragmatic realization that no supplier can fully anticipate changes in weather pattern or the potential locale usage of the fuels. For example, fuels designed for one local set of ambient temperatures may be transported to another geographical region with much colder weather. This point is especially pertinent for the on-highway transportation sector, where trucks can travel long distances, carrying with them fuels suitable for warmer climate but inappropriate for colder temperatures. Even within the same locale for which fuel has been produced, an unseasonal cold snap or extreme swing in temperatures can disrupt fuel operability.
The second point involves transportation and storage. A batch of diesel fuel, after leaving the refinery, would have traveled via pipelines, barges, or long-haul carrier trucks to reach its destination. Upon reaching the fuel terminal or retail fueling station, the fuel is transferred into storage tank for future sale. In both cases, mixing with residual fuels is possible. As a result, the fuel property (for example, the CP) may have been altered to an extent not necessarily matching that whence it is released from the refinery, and may not provide the same amount of cold temperature resistance for which it is designed. That is an additional advantage of refineries using CP as the cold flow release specification, as any such changes may be easily measured.
The final point may appear more abstruse, but is equally relevant, and concerns the precision of a test method. Recall that the diesel fuel, upon final release by the producers, has been quality tested to ensure low temperature performance by any of the CP, CFPP, or LTFT test method. It turns out that each of these methods is associated with a measure of precision, which may be thought of as the uncertainty in the test result. This uncertainty arises due to many factors such as variations in the test equipment, differences in the techniques among human operators (if a manual test is involved), or even in the sample fuel itself. The effects of these variabilities are typically summarized into a single numerical measure called the precision or reproducibility (or error margin). The larger the reproducibility, the greater the variance in the result. As discussed below, the precision of a test method impacts the practical usage of the fuel.
The precision addresses how much two test results may differ, and yet be accepted as being the same (i.e. statistically not significant). For all of the test methods discussed above, ASTM has calculated well-defined reproducibilities obtained in interlaboratory studies. These numbers can be considered as the average precision of the test method and are tabulated in Table 1:
Table 1: Reproducibility of ASTM Low Temperature Property Test for Diesel:
Test Type ASTM Method Number Reproducibility (°C)
CP D 2500 4.0
D 5771 (a) 3.1
D 5772 (a) 3.7
D 5773 (a) 2.6
LTFT D 4539 4.0
CFPP D 6371 2.6 to 5.9 (b)
(a) D5771, 5772, 5773 are automatic CP test methods
(b) the reproducibility for D6371 is found to be linearly dependent on the CFPP value itself. Samples with warmer CFPPs have smaller reproducibilities than those with colder CFPPs.
For instance, the reproducibility for the (manual) CP test method equals 4.0°C. This signifies that two independent CP test results (using the same diesel sample) may differ by up to 4.0°C, and still be considered statistically the same after taking into account of all the different sources of variabilities inherent in the test method. This point is quite noteworthy but often neglected, but it does indicate how test results should be practically interpreted. To cast into an even more concrete scenario, it is perfectly reasonable for a supplier to advertise (and rightly believe) that his fuel has attained a CP value of say, 0°C (assuming he only measures it once), while another independent test may record a value of +4°C, and both would be considered correct. Depending on the particular circumstance, this variation may spell the difference between an operational truck and a plugged filter.
There are at least two implications of interest. First, different types of tests (CP vs. LTFT vs. CFPP) have different precisions, and should be interpreted accordingly. Secondly, there exists automatic CP test methods that are more precise (i.e. smaller reproducibility) than the manual test. It follows that in a practical application and all else being equal, the more precise test type and the most precise test method within that type is likely to yield the least amount of uncertainty and error when a potential cold-weather problem arises. To minimize the probability of customer dissatisfaction, responsible fuel producers have many incentives to use the most precise and fail-safe test method(s) to qualify their fuels.
Note also, that the reproducibilies presented in Table 1 are based on some particular historical (1988 to 1990) test sets. As test methods and fuel types evolve, these numbers may also change. To this end it is helpful to delineate some of the recent precision findings, if only to gain an appreciation of the continual utility of the test results.
In terms of CP precision, ASTM has maintained an interlaboratory cross-check program in which more than 100 worldwide laboratories participate, and the average reproducibility from 1997 to early 2002 for 16 diesel samples is 4.3°C (for D2500), close to that in Table 1. For method D5773, the average reproducibility over the same 16 samples is 2.0°C. For methods D5771 and also D5772, the average reproducibility over the same 16 samples is 2.9°C.
ASTM has also begun testing for CFPP in late 2001. At this point two un-additized samples have had precisions of 3.0 and 4.2°C respectively. More germane to this discussion, however, are the precisions obtained with additized fuels, which after all, is the primary reason for the development of the CFPP test method. In the above-mentioned 1999 CFPP round-robin conducted by CEN, it is their conclusions that the overall reproducibility of 6.7°C is “substantially worse than currently quoted in EN 116” and that “precision differs greatly between one fuel and another” [ref (2)].
The Netherlands-based Institute for Interlaboratory Studies has also reached similar conclusions in their 1999 and 2001 winter gasoil (i.e. diesel) interlaboratory studies [ref (14)]. The 1999 report notes that “serious analytical problems have been observed” in the CFPP test, with a broad range of reported results from -31 to -16°C, an average result of -21.6°C and a large reproducibility of 11.1°C. It concludes by stating that “the calculated reproducibility was not at all in agreement with IP 309”. The 2001 results echo similar sentiments, yielding a broad range of results from -30 to +2°C, an average CFPP of -23.8°C and a reproducibility of 8.6°C. In comparison, the reproducibility of CP for the same two fuels is 2.6 and 3.2°C respectively.
Since LTFT is time-consuming and therefore impractical in practice, there has been no recent large-scale interlaboratory study to assess and update its precision.
It appears then, of the two low temperature tests widely practiced currently, CP is the more precise method, while CFPP’s precision has deteriorated from its quoted number in the test method.
Methods to address Low Temperature Operability available to the End users
For the consumer faced with a cold-weather fuel problem, an array of solutions is also available at his disposal. Dilution with lighter kerosene (also known as #1 grade diesel) in order to lower the CP is often touted as a viable solution. However, this method is costly (as kerosene costs more) and typically has lower lubricity and also lowers the caloric value of the fuel, often leading to degraded fuel economy. Using after-market additives is another possibility. However, both EMA and many truck and engine manufacturers do not recommend the use of additives [ref (15)]. Furthermore, the additizing process frequently requires that the fuel temperature be warm (considerably above its CP) before there are any signs of wax and that the mixing be thorough and intimate, conditions not easily met when the ambient temperature is already low and fuel is already in the tank. On the other hand, it should also be noted that not all vehicle manufacturers discourage the use of aftermarket additives, individual manufacturer [ ref (14)] may recommend the use of aftermarket additive under low ambient temperature conditions.
Other methods do not propose to modify the nature of the fuel, but attempt to prevent it from waxing in the first place. These include the use of fuel tank or filter heaters, the idling of the engine so to maintain fuel circulation, or the seeking of warm shelters. Some have even suggested the constant filling with local fuels, so to increase the probability of a better match of fuel with local weather conditions.
Although each of the above methods has merit, the shortcomings are equally apparent. Use of heaters places a drain on the energy source, whether battery or secondary generator. The practice of idling for the purpose of keeping the fuel warm undoubtedly wastes fuel, and is increasingly frowned upon by regulatory agencies because of emission concern. The availability of warm shelters, needless to say, presents a logistic difficulty. Finally, the use of local fuels necessitates frequent stoppage for refueling, and is by no means a foolproof guarantee that the refill and end blend would meet the requisite requirement for low temperature performance, especially if an unseasonably cold snap takes place.
A Better and more Confident Way to Predict On-the-Road Low Temperature Performance
Perhaps the biggest obstacle facing the end user dealing with cold-fuel issues is not the lack of solutions, but the lack of transparency between actual fuel property and what he knows of it, leading to the most robust and cost-effective choice of solution(s). It is evident from the above discussion that producers and suppliers have expended efforts in specifying fuels that purport to meet the needs of most users most of the time. The goal of fuel producers is to deliver quality products to their customers with minimal risk of cold flow or other quality problems. The challenge is how to empower the user with this information so that his decisions are less desultory suppositions but more informed and confident.
It should also be apparent from the above that of the two practical parameters of assessing cold temperature fuel performance, CP is more robust and precise. The CP measure is far more immune to variations in particular engine configuration, filter geometry, and interactions between fuel and additives under the kinetics of long time exposure, and in this sense it is more attentive to the interest of the vehicle users, recognizing that it can be overly conservative in predicting the performance of certain additized fuels. Therefore, it is but a short logical step to conclude that if the end users were granted knowledge of the CP of their fuels, their ability to tackle cold-fuel problems, were they to arise, would be considerably enhanced. In that sense, this test has value in serving as an “early” warning to fuel cold flow problems.
The nature of waxing in diesel fuels due to cold temperatures and various test methods to measure diesel fuels’ ability to withstand low temperature are illuminated. Of all the indicators available, cloud point represents the most fail-safe and precise way to interpret low temperature performance. As refineries move towards production of lower sulphur diesel products, maintaining the optimum quality and precision in cold flow will also continue to be a high priority.
Stu Porter is Quality Control Co-ordinator for Suncor Energy Products Inc. Over his 20 years at the Sarnia Refinery he has also served roles as Production Scheduler and Refinery Chemist. He has extensive experience in petroleum testing and is Chair of ASTM Task Groups on Cloud Point and Freezing Point. He received a B.Sc. in Chemistry from the University of Western Ontario and he can be reached by email at firstname.lastname@example.org.
Notes and References
Notice: The citing of commercial companies or products in this article does not denote an endorsement by the author. It Is merely used to illustrate the sources of commonly obtainable and representative information in the public domain.
1. Chevron document: “Diesel Fuels Technical Review”, FTR-2, 1998. Available from
1. “Cold Flow properties of Diesel fuels”, CEN/TC-19/WG-14, February 2000. Chairman: W. Doermer, Secretary: G. Brown
2. BP Australia Fuel News: “Understanding CFPP”, issued on Feb 7, 2002 (ADF0207.doc). Available at
3. For example, Fleetguard® advertises filter sizes between 2 to 25 µm, suitable for a host of engines including Cummins, Mack, Catepillar, and DDC in their OptiGuard product brochure: “LT15075.pdf”. Available from
4. “1981 CRC Diesel Fuel Low-Temperature Operability Field Test “, CRC-528
5. J. Chandler and J. Zechman, “Low-Temperature Operability Limits of Late Model Heavy-Duty Diesel Trucks and the Effect Operability Additives and Changes to the Fuel Delivery System Have on Low-Temperature Performance”, SAE technical paper 2000-01-2883.
6. For CP, there are one manual method ASTM D2500 / IP 219 / ISO 3015 / K 2269 and three automatic methods ASTM D5771 / IP 446, ASTM D5772 / IP 444, and ASTM 5773 / IP 445.
For CFPP: ASTM D6371 / IP 309 / EN 116 / K 2288
For LTFT: ASTM D4539
note: IP = Institute of Petroleum (UK), ISO = International Standards Organization, EN = European Norm, K = Japanese standard
7. ASTM D975 “Standard Specification for Diesel Fuel Oil”, ASTM, West Conshohocken, PA, 2002.
8. EN 590 “Automotive fuels – Diesel – Requirements and Test Methods”, CEN, Brussels, 1999.
9. Worldwide Fuel Charter, April 2000. Available from or
Subsequent comments and responses can be found in
10. “Lessons learned from diesel filter problems”, New Zealand Government press release. Available from: http://www.scoop.co.nz/stories/PA0110/S00249.htm
11. New Zealand Petroleum Products Specifications Regulations, 2002, available from http://www.med.govt.nz/ers/oil_pet/fuelquality/2002regs/2002regs-01.html#P560_21331
12. “EMA Consensus Position on Pump Grade Diesel Specification”, available from
13. EMA FQP-1A Recommended Guideline on Diesel Fuel”, downloadable as in ref (11).
14. Institute for Interlaboratory Studies Proficiency Test of Winter Gasoil: October 1999 (report #IIS 99GO1X) and January 2001 (report #IIS 01G1X), Dordrecht, the Netherlands. Available from
15. see, for example, service bulletins from the followings:
(i) Catepillar: “does not generally recommend the use of fuel additives” (in “Catepillar Machine Fluids Recommendations”, document #SEBU6250-11, October 1999, p.15)
(ii) Cummins: “Never put ‘additives’ in your fuel without first reading your manual or checking with your engine manufacturer to make sure adding anything to the fuel won’t void your warranty… be suspicious of claims that seem too good to be true” (in POWERFORUM newsletter, 2000, volume 2 number 4)
(iii) MTU Friedrichshafen: “The effectiveness of the flow improvers is not guaranteed for every fuel. Certainty is only assured after laboratory testing of the filtering capability” (in “Fluids and Lubricants Specifications”, document # A001061/24E, 2001, p.16)
(iv) EMA FQP-1A Recommended Guideline on Diesel Fuel: “Unless specifically recommended by engine manufacturer or discussed with the fuel supplier in advance, … using aftermarket fuel additives is not recommended as an option for meeting the low temperature operability requirement, because of possible incompatibility with other additives already contained in the fuel”. (from ref (12))
16. see Volkswagen Technical Bulletin, December 8, 2000 on “Diesel Fuel Cold Weather Conditioning for driving at below 0°C., from < http://www.mach.mach-nine.com/gti/vwtechcontent/vwpdf/v200003.pdf>