Energy 28 (2003) 15–26www.elsevier.com/locate/energy

Fuel adulteration issues in Greece

S. Kalligeros, F. Zannikos, S. Stournas, E. Lois∗

National Technical University of Athens, Department of Chemical Engineering, Iroon Polytechniou 9,Athens 157 80, Greece

Received 4 December 2001

Abstract

The fuel adulteration problem is associated with environmental pollution, problems with engine perform-ance, and tax losses. Here, results are presented of a survey in leaded gasoline and automotive diesel,obtained from service stations representative of all the oil companies operating in Greece. For this purpose,165 samples of gasoline and 420 samples of automotive diesel were collected from various parts of thecountry during the years 1998, 1999, and 2000. The gasoline samples were subsequently analyzed for theirkey properties and for any adulteration with cheaper unleaded gasoline. Octane number, benzene, olefins,and total aromatics were determined with the mid-IR method, and the sulfur content with a UVF elementalanalyzer. The analysis of the automotive diesel samples concerned some key properties such as the cetaneindex, density, sulfur content, and the distillation properties of the fuel. The results indicate that there isa large fluctuation of fuel properties among the oil marketers. Examination of the quinizarin content (thetracer of unleaded gasoline) has shown that 11 leaded gasoline samples (8.8%) were mixed with unleadedgasoline, 5 leaded gasoline samples (4%) were mixed with aromatic solvents, whereas about 28% of theautomotive diesel samples suffered from some degree of adulteration, mainly with cheaper heating fuel;and one automotive diesel sample was adulterated with a lighter fraction. Fuel misuse is a common problemnot only for European countries but for practically every nation in the world. The European Union recentlyexpressed its concern on this issue, mandating that by the year 2002 all the member states will promotethe development of a uniform system for fuel quality monitoring. 2002 Elsevier Science Ltd. All rights reserved.

1. Introduction

Fuel quality in recent years has become increasingly important, not only for its role in theactual performance of the vehicles, but for its impact on their emissions [1,2]. Fuel composition

∗ Corresponding author. Tel.+30-772-3190; fax:+30-619-7750.E-mail address: elois@orfeas.chemeng.ntua.gr (E. Lois).

0360-5442/03/$ - see front matter 2002 Elsevier Science Ltd. All rights reserved.PII: S0360-5442(02)00091-9

16 S. Kalligeros et al. / Energy 28 (2003) 15–26

will continue to grow in importance as the governments continue their strategy for the targetof near-zero-emission vehicles [3,4]. Over the past two decades the European Community hasprogressively strengthened its emissions standards for most vehicles. Emissions from passengercars meeting the latest standards enforced in 1996/97 will be 90% lower than those required inthe 1970s. The new draft directives on vehicle exhaust standards and fuel quality envisage twostages. Mandatory emission exhaust limit values are proposed for the year 2005, followed bylower levels in specified years [5]. The specifications of the fuels produced by the refineriesusually comply to those of the legislators. However, the fuel pump at the service station is thepoint at which the actual specifications of the fuels should be ascertained. This paper presentsresults of a survey of gasoline and diesel samples obtained from service stations in Greece, wherethe public buys its fuel.

In Greece, two main types of gasoline are sold in the service stations: leaded gasoline with aResearch Octane Number of 96 (96 RON) for the non-catalytic cars, and unleaded gasoline witha Research Octane Number of 95 (95 RON) for newer cars equipped with a catalyst. Some servicestations also sell premium unleaded with a Research Octane Number of 98 (98 RON) but themarket share of this product is very small (below 2% of the gasoline market). Unleaded gasolineis the cheapest gasoline and it is marked with quinizarin, while leaded and premium unleadedgasoline have similar prices (but they are quinizarin-free). This price differential is the mainmotive for mixing the cheaper with the more expensive fuel. Most gasoline adulteration casesinvolve the illegal mixing of the cheaper unleaded into leaded gasoline. Less common is themixing of much cheaper heating fuel into gasoline. In the case of the adulteration of similar fuels(gasoline with gasoline), the main impact is lost taxes and increased emissions, whereas in thecase of mixing dissimilar fuels (heating fuel with gasoline), long-term use may also lead toengine damage.

Three main types of middle distillates are sold in Greece: automotive diesel fuel [6], heatingfuel, and marine fuel. Heating fuel is cheaper than automotive diesel and is colored red. It containsfurfural as a chemical marker at a proportion of 20 mg/lt and its sulfur content is up to 2000ppm [7,8]. Marine fuel is also cheaper than automotive diesel. It is colored black and containsquinizarin as a chemical marker at a proportion of 5 mg/lt. It also has a sulfur content up to10,000 ppm (1% wt) [9]. The large price difference of heating and marine fuel with automotivediesel fuel — due to the different tax policy — is the main incentive for the adulteration. Themost common practice in the illegal market is to remove the color from the heating and marinefuel with clay treatment using discoloring earths and sell them as automotive diesel. Laboratoryexperiments show that the discoloring process completely removes the furfural marker but doesnot affect the sulfur content of the fuel. For this reason the sulfur content can be used as a physicalmarker to characterize the fuel quality.

The limit of sulfur content for automotive diesel was 500 ppm (maximum) until December1999 and 350 ppm from January 1, 2000 [10]. Ultra-low-sulfur (� 5 ppm) diesel fuels have beenintroduced in Sweden and Finland to reduce the emitted particulate levels from urban buses [11–13]. A 32–44% reduction in particulate matter was obtained in buses with pre-Euro I engines,and an approximate 32% reduction in particulate matter was obtained in buses with EURO IIengines, using ultra-low-sulfur European “City” diesel fuel [14]. However, other fuel propertiessuch as density and volatility may also affect emissions [15], but the degree of influence dependsvery much on the engine design [16]. For a number of reasons, fuels produced by the refineries

17S. Kalligeros et al. / Energy 28 (2003) 15–26

usually comply with legislation, but alterations in the fuel properties may occur during transpor-tation and up to the point where the fuel is dispensed into the consumer car tanks. The EuropeanUnion expressed its concern on this issue through European Directive 98/70, Article 8, statingthat the member states will promote the development of a uniform system for fuel quality monitor-ing [17]. By 30 June 2002, every member state shall submit their report for the preceding calendaryear to the Commission.

In this paper, we present the results on the quality of both gasoline and automotive diesel atthe fuel pumps of service stations in Greece, and we examine the adulteration problem encounteredin these fuels for the years 1998–2000.

2. Experimental procedure

One hundred and sixty five samples of gasoline were collected from service stations locatedall over Greece (see map). They consisted of 125 samples of leaded gasoline, 37 samples ofunleaded gasoline, and 3 samples of premium unleaded. Emphasis was given to leaded gasolinebecause of its higher price, which increases the motivation for adulteration. Most of these sampleswere collected during the year 1999 and some during 1998. Almost all unleaded gasoline sampleswere collected during the first months of the year 2000 in order to monitor the degree of responseof the market to the tighter specifications that came into effect from the beginning of the year2000. The following gasoline properties were determined since they are directly related to theexhaust emissions: octane number, benzene, olefins, and total aromatics were established usingthe mid-IR method and sulfur content was measured using an ANTEK 9000NS elemental ana-lyzer[18]. A spectrometer (Petrospec GS-1000 Gasoline Analyzer) offered by Varlen Instruments,was applied for the mid-infrared spectra analysis [19–22]. For calibration purposes we usedrefinery gasoline samples measured with the standard ASTM methods for RON, total aromatics,olefins and lead content. Each value reported here is the average of at least two scans. A recalibrat-ing procedure was performed every thirty scans with simultaneous fuel filter changing. It shouldbe mentioned that the mid-IR method does not record the lead content of leaded gasolines andis completely insensitive and “blind” to the lead anti-knocking additives. For this reason, therewas a difference of about 3–4 RON units between the actual ASTM RON values and the mid-IR measured RON for the leaded gasoline samples. Therefore, the RON values for leaded gasolineas shown in Fig. 1 were recalculated from the mid-IR results using an average lead content of0.15 g/L based on data from the refinery samples. The recalculation procedure was necessary forcomparison reasons and for a more realistic presentation of the RON values of the gasolinesamples in Fig. 1. A quinizarin marker is added to unleaded gasoline at a concentration of 6mg/lt, while leaded and premium unleaded gasoline are quinizarin-free. Gasoline adulteration withmarked unleaded gasolines is easily found from the quinizarin levels. The quantitative analysismethod involves the measurement of the quinizarin absorbance at 605 nm, employing a Jenway6100 spectrophotometer [23,24].

Four hundred and twenty samples of automotive diesel were collected from 1998 to 2000 andthe following properties were determined: sulfur content, cetane Index, density, and distillationparameters, since they represent most concern for the exhaust emissions. The methods of measure-ment are given in Table 1.

18 S. Kalligeros et al. / Energy 28 (2003) 15–26

Fig. 1. Distribution of the octane number for the gasolines.

Table 1Determination Methods for Fuel Properties

Property Method

Research Octane Number Correlates to ASTM D-2699Distillation, (wt. %) ASTM D-86Sulfur content, (ppm) ASTM D-5453Benzene content, (vol. %) ASTM D-6277aOlefins content, (vol. %) ASTM D-1319Aromatics content, (vol. %) ASTM D-1319Cetane Index ASTM D-4737Density ASTM D-1298

3. Results and discussion

3.1. Gasoline

The first stage of this study was devoted to the collection of the samples from service stationsacross the country. Emphasis on the sampling was given on leaded gasoline, because of its higherprice that increases the motivation for adulteration. Examination of the quinizarin tracer has shownthat 11 leaded gasoline samples (8.8%) were mixed with unleaded gasoline. They are marked as

19S. Kalligeros et al. / Energy 28 (2003) 15–26

Fig. 2. Ratio of aromatics/olefins for the gasoline samples.

circles with black background in Figs. 1–5. Most of the samples were collected during the year1999 and some during 1998. In contrast, almost all of the samples of unleaded gasoline werecollected during the first months of the year 2000 in order to monitor the degree of response tothe tighter specifications that came into effect at the beginning of the year.

Fig. 3. Aromatic content for the gasoline samples.

20 S. Kalligeros et al. / Energy 28 (2003) 15–26

Fig. 4. Sulfur content of the gasolines.

Fig. 5. Benzene content of gasolines.

3.1.1. Octane numberFig. 1 depicts the octane number of all the three types of gasoline samples. Generally there

were no problems with the octane number of the gasolines, irrespective of type, given the accuracyof the method (± 1 ON). Only one sample of leaded gasoline was found to be outside the specifi-

21S. Kalligeros et al. / Energy 28 (2003) 15–26

cation limit. The exceptionally high ON of another adulterated leaded gasoline sample (99 RON)was due to the addition of aromatic solvents (mainly toluene and xylene).

3.1.2. Aromatics and olefins contentLeaded gasoline has a high olefin content, since its main components come from fluid catalytic

cracking refinery units. In contrast, the olefin content in unleaded gasolines is relatively low, sincethe main constituents come from catalytic reforming units in order to reach the desired octanenumber in the absence of the lead additives (TEL). The high quantities of reformate used in thegasolines give high concentrations of aromatics in unleaded gasolines — in some cases in excessof the highest allowable limit of 42%, which is the current specification since 1 January 2000.The adulterated samples of leaded gasoline exhibit different behavior from the normal typicalsamples — they have relatively low concentrations of olefins and high concentrations of aromaticconstituents. If we calculate the ratio of aromatics/olefins for all the gasoline samples, the adulter-ated samples of leaded gasoline can easily be traced by comparison with the normal samples (Fig.2). This is supported by the findings of the aromatic content of the gasolines (Fig. 3) where thegasolines with high aromatics/olefin ratios (Fig. 2) exhibit high aromatics concentrations as well.We can say with confidence that if the ratio of aromatics/olefins is greater than two in a leadedgasoline sample, then this gasoline is most likely adulterated, irrespective of the existence of thelegal tracer.

3.1.3. Sulfur contentDue to the high specification limit of sulfur in leaded gasoline (1000 ppm maximum), all the

samples exhibited normal behavior (Fig. 4). The same conclusion was reached for the unleadedgasoline with a much more severe limit (150 ppm since 1 January 2000); most of the samplescollected during the first few months of 2000 were inside the specification.

3.1.4. Benzene contentDue to the increased toxicity of benzene, the maximum allowed concentration for all types of

gasolines was reduced from 4% to 1% vol., as of 1 January 2000. As we have already mentioned,all the samples of leaded gasoline were collected before 31 December 1999, and they all fulfillthe specification that was effective. In contrast, 36 out of 40 samples of unleaded gasoline thatwere collected after 1 January 2000 do not meet the effective specification (Fig. 5) due to thefact that most of the samples were collected during the months of January and February 2000, aperiod in which the refineries were not fully harmonized with the new, more severe, specifications.

3.2. Automotive diesel

Four hundred and twenty samples of automotive diesel were collected during the years 1998,1999, and 2000.

3.2.1. Distillation characteristicsFig. 6 presents the distillation curves of all the samples of automotive diesel. Greece is a country

with a changing climate. The Greek refineries do not blend lighter fractions (e.g., kerosene) intodiesel fuel to face cold flow properties, and they usually apply additives to reduce cloud and pour

22 S. Kalligeros et al. / Energy 28 (2003) 15–26

Fig. 6. Distillation curves of the automotive diesel fuels.

points and to improve winter operation. The samples that were adulterated with lighter fractionsare clearly identified.

3.2.2. Sulfur contentFig. 7 shows the sulfur content of all the automotive diesel samples. The specification was

reduced from 500 ppm to 350 ppm from 1 January 2000. Fig. 7 shows the samples which areoutside the specifications, given the accuracy of the method (ED-XRF in 1998, and UVF in 1999and 2000). About 28% of the samples for all the years was found to be adulterated, mainly withheating fuel.

3.2.3. Cetane number–cetane IndexThe cetane number and the cetane index reflect the ignition quality of the automotive diesel,

since it affects the engine performance, the efficiency, the engine emissions, and the noise pro-duced. The cetane index is not measured directly, but is calculated according to the ASTM D-4737 method from other diesel characteristics (distillation points and density). The cetane indexdid not change in the year 2000 and remained at 46 minimum (Fig. 8). In Greece it is traditionallyconsidered that the cetane index of the various diesels is too high due to the types of crudes andthe nature of the refinery processes used to produce them. However, according to our measure-ments, the average cetane index in 1998 was 52.5, and it was reduced to 51.6 and 50.2 for thesamples in the years 1999 and 2000. This reduction is explained by the fact that in the productionprocess, increasingly greater amounts of fractions coming from the cracking units are used, whichinherently have low cetane index and low cetane number.

23S. Kalligeros et al. / Energy 28 (2003) 15–26

Fig. 7. Sulfur content of the automotive diesel fuels.

Fig. 8. The cetane index of the automotive diesel.

3.2.4. DensityDensity is a significant property because it controls the amount of fuel that is compressed and

burned in the combustion chamber. The higher the amount of fuel sprayed into the combustionchamber, the higher the output of partially oxidized products emitted. The specification for theupper limit was reduced from 0.860 gr/cm3 to 0.845 gr/cm3 on 1 January 2000, when the minimum

24 S. Kalligeros et al. / Energy 28 (2003) 15–26

limit value remained at 0.820 gr/cm3. It can be seen that the average value in 1998 was higherthan in 1999. This was due to the adjustment of the refineries to the new European directive ofdiesel fuel density. One sample in 1998, mixed probably with some lighter fuel, did not complyeven with the current legislation. On the other hand, seven samples in the year 2000 had a densityhigher than the current regulation. This is probably the result of mixing heavier diesel fractionswith automotive diesel fuel. The density of all samples is highlighted in Fig. 9.

3.2.5. Geographic distribution of the adulterated samplesFig. 10 shows the geographic distribution of the normal and the adulterated samples. The areas

with the most adulterated samples are Athens, Thessaloniki, and Thessaly. Increased numbers ofsamples outside the specifications were also found in the Aegean islands, in Ipiros, Thrace, andEvia. In Pelloponisos, Sterea Ellada, and the rest of Macedonia (except Thessaloniki), the percent-age of the samples outside specifications were lower than the national average, but were morethan 10%. From all the areas only Crete and the Ionian islands had the smallest number ofadulterated fuels.

4. Conclusions

Results are presented of a survey of leaded gasoline and automotive diesel, obtained fromservice stations representative of all the oil companies operating in Greece. For this purpose, 165samples of gasolines and 420 samples of automotive diesel were collected from various parts ofthe country during the years 1998, 1999, and 2000. The leaded gasoline samples were sub-sequently analyzed for any adulteration with cheaper unleaded gasoline, and for some of the keyproperties, i.e., octane number, and sulfur, benzene, olefins, and total aromatics content. The

Fig. 9. The density of the automotive diesel.

25S. Kalligeros et al. / Energy 28 (2003) 15–26

Fig. 10. The geographical distribution of the adulterated samples of automotive diesel.

analysis of the diesel samples concerned some key properties of the automotive diesel, such ascetane index, density, sulfur content, and the distillation properties of the fuel. From this studyit was concluded that:

1. There is a large fluctuation for the fuel properties among the retailers.2. Eleven leaded gasoline samples (8.8%) were mixed with unleaded gasoline based on the quin-

izarin tracing.3. Five leaded gasoline samples (4%) were mixed with aromatic solvents.

26 S. Kalligeros et al. / Energy 28 (2003) 15–26

4. Twenty eight percent of the automotive diesel samples were found to be adulterated with heat-ing or marine fuels based on the sulfur concentration.

5. One automotive diesel sample was found to be adulterated with lighter fractions.6. Most of the other key properties of the gasoline and automotive fuels in Greece were found

to comply with the current EU legislation.7. The findings of this research verified the concerns of the European Union for fuel adulteration,

which mandated that by the year 2002 all the member states will promote the development ofa uniform system for fuel quality monitoring.

References

[1] Singal KS, Pundir PB. Diesel Fuel Quality and Particulate Emissions: An Overview. SAE Paper 961185. 1996.[2] Singal KS, Singh PI, Pandey CD, Runda KM, Semwal BP, Gandhi KK. Fuel Quality Requirements for Reduction

of Diesel Emissions. SAE Paper 1999-01-3592. 1999.[3] Colucci JM, Darlington TL, Kahlbaum DF. An Analysis of 1996 Gasoline Quality in the United States, SAE

Paper 982723. 1998.[4] Colucci JM, Darlington TL, Kahlbaum, DF. An Analysis of 1996–98 Gasoline Quality in the United States, SAE

Paper 1999-01-3584. 1999.[5] Federation Internationale De L’ Automobile. 2001, 8 Place de la Concorde, 75008 Paris, France. See also:

http://www.fia.com/tourisme/enviro-a/cleana.htm[6] Greek Government Gazette. Automotive Diesel Fuel Requirements and test methods. 1994; 336 (B): 2984-2986.[7] Greek Government Gazette. Heating Diesel Fuel Requirements and test methods. 1992; 153 (B): 1446.[8] Greek Government Gazette. Marking and Colouring Procedure. 1993; 496 (B): 5431-5432.[9] Greek Government Gazette. Marine Fuel Requirements and test methods. 1999; 133 (B): 2103-2104.

[10] European Standard. Automotive Fuels. Diesel-Requirements and test methods. EN 590. 1999. European Committeefor Standardization. Rue De Stassart 36 B-1050 Brussels.

[11] Saikkonen P, Mikkonen S, Makela M, Niemi A. Lubricity of Reformulated Diesel Fuel — Experience in Finland.SAE Paper 961948. 1996.

[12] Miura M, Ikeda T, Takizawa H, Yoshida H, Ikebe H. Study on Lubricity of Low Sulfur Diesel Fuels. SAE Paper972895. 1997.

[13] Tucker RF, Stradling RJ, Wolveridge PE, Rivers KJ, Unbbens A. The Lubricity of Deeply Hydrogenated DieselFuels — The Swedish Experience. SAE Paper 942016. 1994.

[14] Atkinson CM, Thompson GJ, Traver ML, Clark NN. In-Cylinder Combustion Pressure Characteristics of Fischer–Tropsch and Conventional Diesel fuels in a Heavy Duty CI Engine. SAE Paper 1999-01-1472. 1999.

[15] Gibbs LM. The impact of state air quality and product regulations on current and future fuel properties. In: StraussKH, Dukek WC, editors. The Impact of US Environmental Regulations on Fuel Quality. Philadelphia: AmericanSociety for Testing and Materials; 1992. p. 30–60.

[16] Lee R, Pedley J, Hobbs C. Fuel Quality Impact On Heavy Duty Diesel Emissions: A literature Review. SAEPaper 982649. 1998.

[17] Directive 98/70/EC of the European Parliament and of the Council Relating to the Quality of Petrol and DieselFuels and Amending Council Directive 93/12/EEC. 1998.

[18] ANTEK Instruments Inc. ANTEK 9000NS manual, Nitrogen/Sulfur Analysers. 1998.[19] Swarin SJ, Drumm CA. Prediction of Gasoline Proprieties with Near-Infrared Spectroscopy and Chemiometrics.

SAE Paper 912390, 1991.[20] Fodor GE, Kohl KB. Analysis of Middle Distillate Fuels by Mid Infrared Spectroscopy. Energy and Fuels

1993;7:598–601.[21] Fodor GE. Analysis of Petroleum Fuels by Mindband Infrared Spectroscopy. SAE Paper 941019. 1994.[22] Janis B. Fuel Quality Control by Mid Infrared Spectroscopy. SAE Paper 1999-01-1546. 1999.[23] European Standard. prEN 214. 1986. European Committee for Standardization. Rue De Stassart 36 B-1050 Brussels.[24] Greek Government Gazette. Marking Unleaded Gasoline. 1992; 403 (B): 3787-3788.