Has India’s Ujjwala Program enabled millions to shift to Liquefied Petroleum Gas as a primary cooking fuel?

Abhishek Kar, Devyani Singh, Hisham Zerriffi

The global energy poverty problem associated with cooking with solid fuels such as firewood, crop residue, cattle dung, and coal is well recognized and at the core of Sustainable Development Goal 71. A key solution being promoted is to facilitate households switching to cleaner burning fuels such as Liquefied Petroleum Gas (LPG). India’s Pradhan Mantri Ujjwala Yojana (PMUY; hereafter, Ujjwala) is a flagship program to offset the high upfront costs of LPG for poor households2. It is targeted at poor women who are dependent on unprocessed solid fuels. The government has also expanded the distribution network to reduce supply delays and “last-mile” transport. As a result, 35 million poor households have received a new LPG stove within the last 22 months. Critics argue, however, that Ujjwala beneficiaries still engage in fuel stacking resulting in low LPG refill purchases (measured by the number of cylinders of 14.2 kg LPG purchased).

Despite the widespread criticism of PMUY’s refill trends, our analysis of national-level consumer cylinder purchase data (as in May 2018) suggests an encouraging trend on this front. Out of 3.7 crores PMUY connections released in the last 2 years, 2.1 crore PMUY beneficiaries have been LPG customers for one year or more (as of May 2018).  Analysis of cylinder purchase data of this group of beneficiaries (one year as Ujjwala consumer) indicates 30% of them (63 lakh households) have purchased five or more cylinders in their first year (including the cylinder provided during installation). These 63 lakh poor rural households have transitioned to LPG as their majority cooking fuel within the first year. Most of them would likely not have become LPG consumers without the assistance of PMUY. State wise, Uttar Pradesh, West Bengal and Bihar have the highest numbers of PMUY beneficiaries where 36%, 28% and 39% PMUY beneficiaries have adopted LPG as primary cooking fuel respectively. The national average of refill of these customers is 4.1 (including the installation cylinder).

How much of a household’s cooking energy does this represent? Is it enough to make LPG the primary fuel in these Ujjwala households’ cooking energy mix? These are technical questions that are generally left to energy researchers to debate and discuss. However, as Ujjwala’s refill patterns have garnered the widespread attention of policymakers, researchers, media, it has taken on significant policy (and political) ramifications. To answer this question, we estimate how much useful cooking energy is needed, how much LPG is needed to cater to those needs and examine the implications of these Ujjwala refill trends.

Useful Energy refers to energy delivered to the cooking vessel, which accounts for the energy content of the fuel and net combustion efficiency of stove/fuel combination. Globally, variation in income levels, cooking and diet preferences fuel a wide range for useful cooking energy requirements that can vary by up to an order of magnitude (0.77-7.22 MJ/capita/day)3. Our analysis of India’s energy consumption survey data suggests that high-income households in urban areas using LPG have higher cooking energy consumption levels compared to low-income rural households dependent on firewood. As Ujjwala specifically targets poor households, it may be more prudent to use the consumption data of low-income families in India who use firewood. Table 1 presents a range of estimates specific to the Indian energy context. It includes data from published reports and articles as well as calculations based on solid fuel consumption data from India’s national sample survey 68th round and field survey data from households in rural Karnataka, India1.

Table 1: Range of estimates of useful cooking energy in India (For detailed calculation, assumptions, and global estimates, please refer to the supporting information Table S1)

Original Data Estimate of Useful Energy (MJ/hh/Year) for India


For conversion

7 MJ/hh/day 2450 350 days per year of cooking to account for festivals, travel, etc. where cooking not conducted at home For all households4
1300 kg/hh/year 2808 14.4 MJ/kg of fuelwood (at 20% moisture content); 15% net thermal efficiency (product of combustion and heat transfer efficiencies) for traditional biomass stove For rural households in India exclusively using fuelwood: Extracted from NSSO Round 68(year 2011-12) dataset used for fuelwood displacement assessment due to access to LPG1
4.37 kg/hh/day 3304 14.4 MJ/kg fuelwood, 15% efficiency, 350 cooking days/year Fuelwood only users in rural Karnataka: Actual measured wood consumption as per Kitchen Performance Test protocol
12.13 MJ/hh/day 4246 350 days/ year For all Indian households5
6.2 GJ/capita/day Final Energy 4651 15% efficiency, 5 people per household Rural households3

In face of such diversity, selecting a range of useful cooking energy for the Ujjwala beneficiaries for analysis is a challenging (and controversial) task in the absence of Ujjwala specific field data. Even with sample field data, questions related to its representativeness considering India’s extraordinary diversity in cooking methods will exist. We use three scenarios to capture the range of useful cooking energy requirement annually for Ujjwala households: 2500 MJ/ hh, 3500 MJ/hh, and 4500 MJ/hh. Typically published estimates of useful cooking energy in rural households of other developing countries are also within this range3.

Commercial LPG stoves get one to five stars based on efficiencies tiers that vary from 68% to above 81%. We assume that Ujjwala beneficiaries would use cheaper burners with lowest LPG stove tier (mean efficiency of 70%). However, often rated efficiencies are not representative of field conditions or they decline over time, so we assume 10% reduction over the rated efficiency levels with effective stove efficiency of ~60%3. As LPG’s calorific value is 45.84 MJ/kg, the useful energy delivered per cylinder would be 385 MJ.

If an Ujjwala household with cooking energy needs defined in the three scenarios above were to switch all of its cooking over to LPG the number of cylinders required would range between 6.5 to 11.7 per year. The mid-range estimate would be 9.1 cylinders per year (3500 MJ of energy requirement/ 385 MJ of energy delivered per cylinder). What does this imply for that one-third of Ujjwala customers (>10 million households) purchasing 4 or more cylinders per year (including the initial cylinder)? At a minimum, it means anywhere from 34% to 62% (44% for mid-point estimate) of their energy is being met by LPG assuming just 4 cylinders. For households consuming 5 cylinders, it would cater to 43% – 77% of energy requirement (55% at midpoint estimate). Hence, households at the lower (2500 MJ-3500 MJ) and higher end (3501 MJ- 4500 MJ) of the useful energy spectrum make LPG the primary (>50%) cooking fuel with four and five LPG cylinders per year respectively.

This suggests the possibility of a remarkable energy transition. One in every three Ujjwala beneficiary adopted a commercial fuel like LPG as the primary cooking fuel or is very near to this cut-off point without any change to their income levels or changes in access to non-monetized solid fuels. LPG is now the main fuel of their energy mix in a relatively short period of time, a stark departure from their past where solid fuels would have been their exclusive or primary cooking fuel. It suggests that access is not only the starting point of transition but also a driver by itself towards greater use of clean fuels for millions of poor households.

However, primary LPG users (³ 4/5 cylinders per year or even more) account for only 43% of Ujjwala beneficiaries. The majority are more occasional users (37%) using 2-3 cylinders annually or rare users (20%) who have not come back for refills. Furthermore, for most “primary” users, solid fuels are still used in traditional stoves. Such fuel stacking limits the health benefits accrued from LPG usage2. Additional measures to change behaviors are required in the long run to move households towards exclusive or near-exclusive usage of clean fuels2.

For Supplemental Information please see: Ujjwala Refill SI Tables


(1)      Singh, D.; Pachauri, S.; Zerriffi, H. Environmental Payoffs of LPG Cooking in India. Environ. Res. Lett. 2017.

(2)      Kar, A.; Zerriffi, H. From Cookstove Acquisition to Cooking Transition: Framing the Behavioural Aspects of Cookstove Interventions. Energy Res. Soc. Sci. 2018, 42, 22–33.

(3)      Daioglou, V.; van Ruijven, B. J.; van Vuuren, D. P. Model Projections for Household Energy Use in Developing Countries. Energy 2012, 37 (1), 601–615.

(4)      Niti Ayog. India Energy Security Scenarios 2047 Version 2.0. Gov. India 2015.

(5)      Purohit, P.; Kumar, A.; Rana, S.; Kandpal, T. C. Using Renewable Energy Technologies for Domestic Cooking in India: A Methodology for Potential Estimation. Renew. Energy 2002, 26 (2), 235–246.



An earlier version of this article (3 April 2018) cited two figures presented by the Ministry of Petroleum and Natural Gas (MoPNG) which were publicly available. First, 80.2% of Ujjwala beneficiary households have come back for a refill within the first year (they get the initial cylinder at the time of connection). Second, 47.1% of these returning customers have purchased four or more cylinders (including the initial cylinder) in one year (one-third of all Ujjwala beneficiaries).

The revised numbers of MoPNG in the updated version of this article is  more than the earlier number for reasons detailed below:
The earlier estimate of refill pattern by the ministry of petroleum and natural gas (MoPNG):
1) Identify all customers who have got four or more refills in a one year period (say, Jan- December 2017) (A).
2) Calculate all who became Ujjwala beneficiaries within this period (B).
3) % of active customers (who came back for refill) who got four or more refills: A / B: 47.1%
In the revised methodology:
1) Identify all customers who completed at least one year as Ujjwala beneficiaries (365 days from the connection data (Subscription Voucher or SV date) – D
2) Identify all customers out of this pool who purchased 4 or more refills with 365 days of getting a new connection.- E
3) % of active customers (who came back for refill) who got four or more refills: E / D: 54.1%
It has led to an increase of 7% in the proportion of active customers who have purchased 4 or more refills in one year