Diesel hides behind CNG

What’s going on? First the key partners of the Central Pollution Control Board -- IOC, and NEERI -- involved with yet to be released source apportionment study made claims publicly that LPG is the most polluting fuel in our cities. Now in quick succession a second study follows from CPCB that ranks CNG as the “worst” fuel and Euro II-III diesel as the “best”. No other government in the world has every branded CNG as worse than Euro II-III diesel.

This has emerged from the joint study of the Central Pollution Control Board and ICAT called-- Study of the Exhaust Gases from different fuel based vehicles for Carbonyls and Methane Emissions. This study has tested all combination of vehicle and fuels (diesel, petrol, LPG, CNG, bio-diesel and ethanol) for a different genre of air toxics called carbonyl and aldehydes that are normally not regulated in vehicular exhaust. These are known as toxics as they cause harm even at small concentration – ranging from eye irritation to cancer. But the study has also included methane with this bunch which is non-toxic but a greenhouse gas. This odd mix of air toxics that harm public health and methane which is a greenhouse gas has been bunched to rank the fuels.

It is important to generate data on air toxics to guide policy action. But the design, scope and conclusion of the study have raised many doubts about the objective of the study. The study which otherwise has generated valuable data on a range of pollutants from different combination of fuels and vehicles – diesel, petrol, CNG, LPG, bio-diesel, ethanol, has become pawn in the hands of diesel business that is out to use this to discredit CNG. The reasons are the missing links, critical omissions, and discrepancies in the study.

   Why is the study so misleading?
1.  What is the rationale for selecting the odd combination of carbonyl and methane to rank fuels? Why has it left out more toxic and important cancer causing compounds that other governments prioritise?
2.  Why only one greenhouse gas which is typically associated with CNG selected to compare fuels?
3.  Why has the study compared old retrofitted CNG and LPG vehicles with OEM vehicles on diesel and petrol?
4.  Why has the study not mentioned the actual composition of the fuels used for testing the vehicles, especially when the study takes note of the fact that changes in the fuel parameters have bearing on the unregulated toxic emissions?
5.   Wrong and contradictory conclusions drawn in the study
6.   What needs to be done?


1. What is the rationale for selecting the odd combination of carbonyl and methane to rank fuels? Why has it left out more toxic and important cancer causing compounds that other governments prioritise?

•    These very selectively chosen pollutants have been made the basis of the comparison with diesel and other fuels. While there is no harm on generating data in these emissions use of these for ranking of fuels is very misleading. Why has the study ignored more harmful air toxics with much stronger cancer potency for comparison? There are numerous toxic compounds in vehicle exhaust. But globally governments use unit risk factors to prioritise these toxics. Unit risk factors indicates the number of excess cancer cases per million people per microgramme per cubic meter concentration of toxic air contaminants exposed over a 70 year period. These are assigned to each toxic to indicate their relative cancer risks.

This study has omitted to indicate any unit risk factors for any of the carbonyl compounds selected for the study to indicate their relative risks compared to the full range of toxics that other governments monitor and control from vehicles. Other groups of toxics like 1,3-Butadiene and benzene, polycyclic aromatic hydrocarbons, and diesel particulate matter are several times more toxic than the carbonyl group selected by the study. Without the full information on all key toxics ranking of the fuels as this study has tried can give a very misleading policy message.

For instance, the two key toxics in this study that matter the most are formaldehyde and acetaldehyde. These are harmful emissions.  As the table below shows formaldehyde is more harmful than acetaldehyde. But in the overall group of toxics the toxicity of diesel particulate matter, benzene, and 1,3-Butadiene are several times higher than the acetaldehyde and formaldehyde. Acetaldehyde causes 2.7 excess cancer cases per million people per micrgramme per cubic meter of concentration over a 70 year lifetime exposure. But diesel particulate matter cause 300 excess cancer cases and 1,3-Butadiene cause 170. Therefore, if only on the basis of acetaldehyde and formaldehyde CPCB brands conventional diesel (the outdated Euro II and Euro III) as the best and retrofitted CNG as the worst, it will give a very misleading and irresponsible policy message. Would CPCB champion Euro II and III diesel as cleaner and safer than CNG? Will that be scientifically tenable given the known toxicity of diesel particulate and other toxic hydrocarbons?

In fact, an earlier ARAI-MOEF-CPCB study has generated valuable emissions factors for the key groups of air toxics for all vehicle types – aldehydes (formaldehydes and acetaldehydes), benzene, 1,3-Butadiene, and PAHs.  But the new study of CPCB has not taken note of these existing data set instead it has resorted to erroneous ranking without any clear caveat on the toxicity of diesel particulates that are classified internationally as toxic air contaminant and as probable human carcinogen and other more potent toxics. The study must avoid this kind of misleading ranking.

Table: Toxic Air contaminant (Unit Risk Factors)

Toxic Air Contaminant Unit Risk/Million People Detection limit (ppb)
Acetaldehyde 2.7 0.10
Benzene 29 0.05
1,3-Butadiene 170 0.04
Carbon Tetrachloride 42 0.02
Chromium, Hexavalent 150,000 0.06 (in nanogram)
Para-Dichlorobenzene 11 0.30
Formaldehyde 6 0.10
Methylene Chloride 1 0.10
Perchloroethylene 5.9 0.01
Diesel particulate matter 300 N/A


Note: Unit Risk represents the number of excess cancer cases per million people per micrgramme per cubic meter TAC concentration over a 70 year lifetime exposure
A diesel particulate matter unit risk value of 300 is used as a reasonable estimate in the “Risk Reduction Plan to reduce Particulae Matter Emissions from Diesel Fuelled Engines an vehicles (ARB, October 2000)
Source:  California Air Resource Board

2. Why only one greenhouse gas which is typically associated with CNG selected to compare fuels?  

Is this study trying to assess the global warming potential of different fuels or just show that CNG emits more methane without any context for such assessment? If this study was serious about comparing the global warming potential of the fuels then it would have considered all the prominent warming gases especially the most dominant one carbon dioxide and also N2O, black carbon etc. This clearly breeds suspicion about the motive of the study. Only generating data on methane indicates nothing about the comparative warming potential unless related to the quantum of all key warming gases from the different vehicle-fuel systems.

If the objective of the study is to compare fuels then it should have done a lifecycle analysis of upstream and downstream emissions as is done globally.

3. Why has the study compared old retrofitted CNG and LPG vehicles with OEM vehicles on diesel and petrol? It is very strange that this study published in 2010 has chosen to ignore the OEM vehicles on CNG and LPG and have considered only retrofitted vehicles. Both bus and car manufacturers are now producing CNG and LPG models. Nearly all city bus projects are based on new dedicated models. Delhi has already taken a decision not to retrofit buses. Also OEM car models on gaseous fuels have multiplied in the recent years. Yet the study states, “For Compressed Natural Gas (CNG), retro-fitment engines have been used due to unavailability of dedicated engine in the market.” (Pg 43). This is inexplicable.

CPCB has justified in a media report that the reason for selecting retrofitted vehicles for the study is the dominance of retrofitted cars in the Indian market. If dominance and availability in the market are the key criteria then by that logic why has the study used diesel fuel with 10 ppm sulphur content to test Euro III diesel vehicles when 10 ppm sulphur diesel is not available in the Indian market? Indian diesel currently has 350 ppm sulphur and ony 13 cities have 50 ppm sulphur diesel. .

Why couldn’t this study also test OEM CNG vehicles in addition to the retrofitted vehicles? The study has to give credible explanation for this.

The study has therefore ended up comparing apples with oranges when they are not directly comparable. This is misleading. You may recall the case from California where the California Air Resources Board had first carried out tests and compared state of the art diesel buses fitted with CRT etc with conventional CNG buses without cat converters and concluded that CNG buses emit more carbonyl and aldehydes than diesel buses. The study was repeated on CNG buses fitted with after treatment systems and they found significant reduction in the toxic emissions.  

Nonetheless, emissions data of retrofitted vehicles are also important to guide policy action on retrofitment. Retrofitment – if not done well, wear and tear of the old vehicles, often absence of proper cat converters – can lead to unstable and high gaseous emissions. The study should have given proper policy guidance on the merit of promoting dedicated vehicles, and enforcing stringent quality audits for proper retrofitment and in-use monitoring. The study has not done that.

Moreover, the CPCB-ICAT study is completely silent on the vintage and age of the retrofitted vehicles selected for the tests. Nor has it given any detail regarding the presence or the status of the aftertreatment devices in the vehicles. This is a serious lapse as that would determine the emissions as well as comparability. It is misleading that a CRDI powered Euro III diesel car fitted with oxicat is compared with a retrofitted CNG car of unknown vintage and model year to conclude that diesel is best and CNG is worst -- and that too without any caveat to diesel particulate emissions.

The study should be modified to separate out the retrofitment vehicles and not club them with OEM vehicles for direct comparison. Additionally, data should be generated for OEM vehicles as well.

4. Why has the study not mentioned the actual composition of the fuels used for testing the vehicles, especially when the study takes note of the fact that changes in the fuel parameters have bearing on the unregulated toxic emissions?


The study shows that the toxic emissions increase in petrol cars from BSII to BS III level. The study explains this on the basis of literature survey that attributes the increase to the reduction in aromatics, olefins etc in petrol fuel. Similarly, improvement in cetane number and reduction in sulphur levels (reference fuels as per regulations have very low sulphur) have been cited as the reasons for increase in emissions from BSII to BSIII diesel cars.

But the study has not analysed the fuel properties of the fuels used. The study only mentions specifications of reference fuels which is not adequate. There is no information on actual oxygen or aromatic content of the fuels used for testing. How can it then correlate the changes with the changes in fuel composition? Globally studies have been carried out on the influence of oxygen, aromatics, olefins and sulphur on aldehydes emissions. And these studies show that it might be possible that aldehydes are more dependent on oxygenates than on aromatics or sulphur content. Studies show oxygen content can also be a very critical parameter regarding aldehydes. Higher aromatics content can also lead to some increase in aldehydes. Sulphur may not show any significant influence. In fact global studies show that higher sulphur content can typically increase emissions from petrol cars due to harmful effect on the three-way catalysts. But the study has not explained any of this. It has also not drawn any policy conclusion from this aspect of their result or suggested any control strategy.

5. Wrong and contradictory conclusions drawn in the study
In some parts of the document the substantive conclusion and the data do not match. For instance, the pai charts on vehicle wise emissions data as well as the comprehensive table on emissions data in the Annexure show clearly the emissions values of various aldehyde and carbonyl compounds from diesel cars. But the conclusion section of the study mentions “Considering Diesel engine fuels, Propionaldehyde, acetone, Crotonaldehyde, Methyl ethyl ketone, n-Butyraldehyde, Methacrolein, Benzaldehyde, m-Tolualdehyde and Hexanal emissions were absent……...”  The same paragraph further mentions – “The aldehyde emissions are absent….” In diesel cars (Pg 61). What can be the reason for this observation and incongruity when the data is showing clearly the emissions level from diesel cars?

6. What needs to be done?

We would urge CPCB to recall and urgently review this study and modify to address the concerns raised and make this study more scientifically tenable. For this review we would like to suggest the following issues for consideration:  

1. The study must explain the rationale for selecting only carbonyls, aldehydes and the reason for combining them with methane. The study should be broadened in its scope to assess emissions of all major air toxics associated with vehicles to avoid misleading ranking of fuels. The study must also indicate the unit risk factors of all major mobile source air toxics and include a caveat on the toxicity of diesel particulates that are classified internationally as toxic air contaminant and as probable human carcinogen

2. Methane data should be separate from the air toxics.

3. If one of the objectives of this study is to compare the warming potential of the fuel-vehicle systems then the study should provide CO2 emissions data and also generate data on other prominent greenhouse gas emissions.  Only data on methane will not help to make the comparison on warming potential of different vehicle-fuel systems. Moreover, the study towards the end has ranked only fuels. But comparison of fuels is normally done based on lifecycle analysis of upstream and downstream emissions.

4. The study should provide the details of the actual composition of the fuels used for testing of the vehicles to explain the impact of the changing fuel parameters on the unregulated emissions. Only mentioning fuel specification is not adequate to explain the impact of changes in fuel parameters on unregulated emissions. It also does not help as the study says that with improvement in fuel quality like lower aromatics and sulphur toxic emissions from petrol vehicles increase. This needs to be explained properly with presentation of the proper data on fuel properties etc.

5. Retrofitted CNG and LPG vehicles should be treated separately in the study and not clubbed and directly compared with OEM vehicles of petrol and diesel. Also take steps to generate emissions data for new OEM CNG and LPG vehicles. Explain the comparability of vehicles by providing the technical details in terms of their vintage, level of emissions control systems etc of the models tested, quality of fuel used.

6. Also review the variance with other emissions data on toxics such as those generated by the Automotive Research Association of India (ARAI). Even though different test data are expected to vary given the type of vehicles tested and test procedures but the variance in this case is huge For instance, the ARAI emissions factor for formaldehyde is 27 times higher than CPCB emissions data for BSIII diesel car. Similar variation is seen for petrol and also retrofitted CNG and LPG cars.  These emissions are also expected to be unstable and variable.

7. This study should also help to identify clear and meaningful policy recommendations on the unregulated emissions from vehicles. This is currently very weak. The study needs to recommend:

•    More stringent emissions standards along with stronger durability requirements and in-use compliance regulations for vehicles in the post 2010 roadmap: CPCB should recommend early nation-wide introduction of Euro V and Euro VI standards to address both regulated and unregulated emissions. Globally the thrust is on improving the efficiency and durability of the after treatment system to cut both regulated and unregulated emissions from vehicles. But this study has completely ignored the role of emissions control systems in different vehicle segments. Only in the case of diesel cars the study mentions, “The aldehyde emissions are absent due to the improved vehicle catalytic converter technologies, location of catalytic converter and better fuel injection technologies.”  (Pg 61). But this aspect has not been highlighted as part of the control strategies for other vehicle categories.  

•    CPCB should also recommend formaldehyde emissions standards for vehicles in the post 2010 emissions standards roadmap. This is the logical conclusion of the study.

•    CPCB should prioritise monitoring of the key air toxics in the ambient air for proper risk management.

The study needs to provide more substantive rationale for its objective, method scope to avoid superficial conclusion. Otherwise it will lead to questions about use of science to make unfair comparison to protect conventional diesel market. It is very worrying that CPCB is projecting Euro II/III diesel as the “best” and not advocating urgent introduction of clean diesel nationwide which is diesel with 10 ppm sulphur content that is used with advanced after treatment systems. At the same time it is also downgrading CNG. The study seeks to perpetrate the status quo. This can enhance health risk in our cities.  

CPCB must address these questions urgently. While it is very important to generate data on unregulated emissions from vehicles for policy action and guidance it is very important to scientifically vet the study as well to avoid misleading conclusions. Erroneous and biased use of such studies can compromise solutions in our polluted cities.