Factors affecting Capacity Utilization of Thermal Power (Coal) Plants in India

As on 31.03.2020, 55.4 % (205135 MW) of total installed capacity (370106 MW) in India is through coal and lignite based power plants. These plants, set up by central, state and private utilities with substantial capital investment are facing consistently reducing Plant Utilization Factor (known as Plant Load Factor, PLF, in India). In the year 201920 the national average thermal power PLF stood at 55.4%, down from 78.6 % in 2007-08. On the other hand, the electricity demand is consistently rising in the country and there exists a peak and energy shortage at national level. In 2019-20 energy shortage was 0.7 % and peak shortage was 0.5 %. A disturbing paradox therefore exists here. On one hand, the country is power deficit, and on the other hand, a large amount of coal based affordable power, ready to be generated by thermal power generators, remains grossly unused. Looking into the fact that considerable investment has gone into developing these thermal power generation assets in the country, the falling PLF is a matter of concern for all the key stakeholders including the power producers, lenders, regulators and consumers. This paper identifies seven major factors that are affecting PLF of thermal power plants and then makes an attempt to project future scenario of PLF so that critical stakeholders can intervene through appropriate actions. Primary research with responses from power professionals has been used to find out the major factors. Future projection of PLF has been done using Partial Least Square (PLS) regression. Projection shows that in the Business As Usual case (Factors increasing at the current CAGR rate), the thermal power plants will face very low level of PLF (14.76 %) by 2024-25. This will mean that many plants will be shut down and many will run for only few hours in a day that too at very low loads. If the future generation mix is kept as indicated by Central Electricity Authority (CEA), a Govt. of India in its report (Draft report on optimal generation capacity mix for 2029-30CEAGovt of India) then the thermal power plant average PLF can sustain above 68 % until 2024-25. If followed, this path can be a breather for the thermal power plants.


Impressive growth of thermal power installed capacity in India
Indian power sector has seen an impressive growth in installed power capacity since independence of the country.
Total installed capacity rose from nearly 1000 MW (as on 31st March 1947) to 370106 MW (as on 31.03.2020). The two tables below (Table and 02) depict the current fuel wise breakup of installed capacity (as on 31.03.2020), (Table   01) and how installed capacity has grown over years (Table 02)-

Growth of power generation capacity in India 1947-2020
Following table (Table 02)    It is also pertinent to note that this drop in Plant Utilization Factor or PLF of coal stations is happening across the sectors (Central, State and Private sector power plants) which is evident from the following table (Table 03). It can also be seen that central sector plants have maintained highest PLF, followed by Private and State sector. On average basis (taking data of 11 years as below), Central sector plants have maintained higher PLF than national by about 10.08 %, the State sector is lower than national PLF by 5.55 % and Private sector is lower than national PLF by 2.13 %. Energy and Peak Deficit in India -Source-Government of India, Ministry of Power and Central Electricity Authority (CEA) websites [3] Under the above background of energy and peak shortages the falling PLF (Table 03) presents a paradox which is the subject matter of work in this paper.

Business Problem
A recent phenomenon called Un-Requisitioned Surplus (URS) indicates the business problem vividly. When a power procuring utility does not requisition power that it had originally contracted to procure from a generating station -it results in the plant running on low PLF thus resulting in unsold electricity for the generator which is termed as Un-Requisitioned Surplus (URS). This is the electricity, power producer is ready to generate but could not generate because the buyer did not requisition it. In 2019-20, in one year alone, the largest power producer of the country NTPC could not generate and sell more than 74000 MUs (URS) [6] of electricity whereas in the same year 0.5 % peak energy deficit and -0.7 % peak deficit was reported at the national level. Demand and supply data given in The Declared Capacity (DC) of a power plant is the capacity in percentage terms at which the plant is ready to generate power. DC is declared by the power producing plant so that the power purchaser can send its requisition accordingly. Such requisition is called the Schedule. If the purchaser does not buy the full Declared Capacity (DC) of the plant it will result in the plant running at lower than the Declared Capacity (DC), hence the Utilization Factor (PLF) shall be lower than the DC. This difference creates the Un-Requisitioned Surplus (URS). [6] The  remunerates the power producer for the cost of fuel burnt) is based on presumption that the plant is running at this specified efficiency level (Normative level). If the producer operates at efficiency levels worse than this normative limit, it could incur losses on this account.
When plants run on low PLF, the efficiency level may drop below this normative level. This means that plant will run with efficiency worse than normative (Heat Rate higher than normative). This, inter alia, means that the producer will spend more on fuel but will get less through energy charge of electricity. Every unit sold in this situation will be at loss. Realising this difficulty, at the representation of thermal power generators, the regulator has permitted some allowance in the normative Heat Rate so as to compensate the generators for such loss. However, the allowances are still not adequate as the PLF is going down day by day and in many power plants the Heat Rate goes worse than normative Heat Rate.
In this context, it is also pertinent to have a look at the way electricity is purchased/sold in the bulk electricity market Kashish Shah, energy analyst with IEEFA (Power Engineering International, 2020) [9] reports that there is already a slowdown in new capacity addition. In the last 12 months to January 2020, 46 GW of coal-fired power projects were formally or informally cancelled. However, 37 GW still under construction. Under this already tight situation for thermal power generation, around 37 GW of new thermal power plants are further in the pipeline (As on March 23, 2020). These pipeline projects, along with fast upcoming new renewable plants, which get presence in scheduling for obvious advantage of being green energy, will push thermal generators to even more difficult situation at a time when demand growth in the country is not very robust.
ET, Dec 20, 2016 [10] report states that if average PLF falls below 48% by 2022 then many coal-based projects may also run into financial difficulty. These projects are financed mainly (75-80%) through loans. Interest on loans must be paid even if plants remain idle. This can jeopardise many projects, hugely burdening the developers and lenders.
This situation requires that we urgently find and focus on the factors responsible for falling Utilization Factor (PLF) of thermal power stations, so that policy makers, power producers, power purchasing companies, consultants and other key stakeholders can take appropriate action to manage the situation optimally. In this paper we explore various factors responsible for such situation and future scenario.

Literature Review
Literature review consisted of a study of published papers, articles, including those published in Energies, Govt.
policy documents, national and international journals, journal websites and websites of Ministry of Power, CEA, NTPC Data, National Electricity Policy, Regional Load Dispatch Center, Ministry of New and Renewable Energy, Conference Proceedings and Presentations etc. ProQuest was used to access the scholarly articles and papers.
Substantial work has been in the area of renewable energy integration in the grid and its consequent effect on thermal power. Papers in this arena have dealt mainly with merits of renewable energy and have looked into how these could be integrated into the grid and how this integration will impact coal plants.
One of the most relevant work in this field comes from Wang, P., & Li, M. (2019) [11]. dissolving overcapacity of thermal power generation and a necessary interprovincial coordination will promote carbon emission reduction rather than investing in coal-fired power plants, and the power authority should turn to alternative investment in cleaner power generation technologies.
Another important work comes from Germany. Robert, K. S. (2019) [13] has analysed hourly generation patterns at large coal-fired units and implications of transitioning from baseload to load-following electricity supplier. The paper states that several factors have led to the decline of electricity generation from coal over the past decade, and projections forecast high rates of growth for wind and solar technologies in coming years. Robert's analysis uses hourly generation data from large coal-fired power stations to determine how operations have been modified in recent years and describes the implications of these changes for plant equipment and unit reliability. The data shows increasing variability in intraday generation output and the high degree of potential impact on coal plant equipment.
The study suggests the development of a new modelling tool that will represent the costs of running coal-  [16] have analysed the financial sustainability of Indian power sector, particularly in terms of poor financial health of the electricity distribution companies in India. They have discussed the poor state of financial health of the distribution companies and the mounting losses that these entities are suffering. Further, they state that distribution sector is revenue generating link which makes it essential to convert these loss incurring utilities into profit ones so that it can cope up with rising energy demand.
Singh, S. K., Bajpai, V. K., & Garg, T. K. (2013) [17] have studied the changes in productivity for 25 state-  Greening The Grid [19] is an extensive study covering operational challenges and cost saving opportunities for renewable energy using state-of-the-art power system planning tools. This paper discusses in detail the integration of renewable and also discusses flexibilisation requirement of thermal power plants in India.
There are many articles from prestigious news publications in India which have cited that the Utilization Factor (PLF) of thermal power plants in India is going down and likely to drop further. They have also brought out some of the factors behind such drop in Utilization Factor.
The Hindu Business Line ( 2017) [20], prestigious news publication in India reports that the coal 'supply constraint' has affected are thermal power plants, which are facing severe coal shortage and are running at less than half-a-day's stocks.  [24] says that all coal-based thermal power plants need to brace for a drastic fall in Capacity Utilization to as low as 48% by 2022, as additional non-thermal electricity generation capacities come on stream. It warns that at that level of utilisation, they may lose the ability to run at a technically viable level and might find it extremely difficult to service debts turning into non-preforming assets for lenders. (2) Exploratory Research through discussion with experts -The domains/areas of concern explored through Literature Review were further discussed with selected experts (10 in total, having work/academic experience of more than 20 years each ) to crystallise to focussed clusters with clearly articulated factors in each cluster which can be tested through a questionnaire.
The above two steps were thought necessary because running a questionnaire without mentioning the factors was not  factor has major impact, then we infer that this is majority opinion and is not a mere chance of random response. We can then reject null hypothesis that p=0.5 and decide that there is conclusive evidence of majority opinion about a particular factor having major impact. This analysis has been done for responses against each question. If 67 % or more have opined that a particular factor has High or very High impact, we have shortlisted that factor as having Major Factor.   to generate when nobody else is able to generate and then back down when others are available.

Regrouping of the factors
The above ten factors have been further regrouped together based on their similarity in terms of outlook and remedial actions and thus we arrive at following seven major factors-   There is no missing data in the set. Since there are three independent variables in our study, we need at least 30 data sets to run regression (Peter's rule of thumb). Here we have taken 35 data sets.

Regression Analysis
The regression was first tried with Ordinary Least Square using Excel but we have later shifted to Partial Least Sqaure (PLS) method using R. Reasons for choosing this method of regression method is as below.
The regression was first started with Ordinary Least Squares (using excel software) and the coefficients of Regression result is as below (Table 8) However, when the data set was tested for multicollinearity, it was found that the X variables had high correlation, despite all the X variables being individually significant. Since the main aim of this part study is the prediction of PLF% in the upcoming years and multicollinearity could have been ignored in these circumstances.
Yet, to make the model more robust, it was decided to use Partial Least Square (PLS) method using R. This method helps reduce the number of predictors to a smaller set of uncorrelated components while retaining the impact of all of the variables on Y. The PLS method squares regression-modelling works by attempting to maximise the covariance in orthogonal space between x scores (T components) and y scores (U components), rather than the variables themselves. Consequently, it does not suffer from the same assumptions concerning data structure as multiple linear regression modelling, and crucially is not invalidated by covariance in predictor variables. It is therefore well suited to time-series data, which is typically characterised by high degrees of covariance amongst predictor variables.

In our PLS model, 2 components were chosen based on RMSEP(Root Mean Square Error of Prediction) criterion
where the number of components for which RMSEP was minimized is chosen. The results was verified by using SelectNcomp function in R which suggests the optimal number of components.
Results of the regression through PLS using R are given below-Cross-validated using 10 random segments. In the present PLS model, for k fold cross validation the data is divided into k equally sized segments, which are usually referred to as folds. Further, k iterations of training and validation are performed so that for each iteration, one fold of data is removed for validation whilst the remaining k-1 folds are used for training the model. The RMSEP is then the mean prediction error for each fold of data removed during cross validation. Consequently, the numbers of components that give the lowest RMSEP are considered optimal for the model (Salter 2018) [26] A 10-fold cross validation is most common and is the approach used in this study to select the two-component Jackknife variance estimator is defined as (g-1)/g ∑_{i=1}^g(\tildeβ_{-i} -\barβ)^2, where g is the number of segments, \tildeβ_{-i} is the estimated coefficient when segment i is left out (called the Jackknife replicates), and \barβ is the mean of the \tildeβ_{-i}. The most common case is delete-one Jackknife, with g = n, the number of observations.
Following result (Table 09) is obtained after using Jackknife test.

Future PLF Projections -Scenario Building & What If Analysis
Using the Partial Least Square (PLS) regression (using R), the Utilization Factor (PLF) has been projected for next

Generation and Demand Mix suggested by Central Electricity Authority vide Draft report on optimal generation capacity mix for 2029-30 (CEAS) [27] (v) Scenario-V-Generation and Demand Mix suggested by Central Electricity
Authority vide Draft report on optimal generation capacity mix for 2029-30 -CEA [27]-Govt of India and assuming that 5000 MW old capacity will be phased out every year. (CEAS+PHOUT).
The national PLF obtained using PLS regression has been bifurcated to Central, State and Private by adding the average historical difference that has been maintained in past between national PLF and these segments (    Predicted Values Scenario -LCAR -Source-PLS Regression Analysis

4) Scenario IV-CEA Projection (CEAS)-Based on Fuel Mix suggested by Central Electricity Authority vide
Draft report on optimal generation capacity mix for 2029-30-CEA-Govt of India)

Scenario-V-CEA based Projection and phasing out of old capacity (CEA+PHOUT) Based on Generation and Demand Mix suggested by Central Electricity Authority vide Draft report on optimal generation
capacity mix for 2029-30-CEA-Govt of India and assuming that 5000 MW old capacity will be phased out every year. (CEAS+PHOUT)

Scenario-III-Low Coal and Aggressive Renewable (LCAR) -Growth of Thermal Power Capacity reducing by 5 % and Renewable Power Capacity growth rate increasing by 2 % (as compared to last 5 years' CAGR)
In this scenario the PLF drops at a reduced rate. National average PLF is projected at 27.01 % in 2024-25. Central sector will maintain 37.10 % PLF and state and private sector will maintain 21.46 % and 24.88 % respectively.
All the above three scenarios show a grim picture. It is pertinent to note that lower PLF not only results in underutilisation of costly assets, it also means that the high efficiency new supercritical power plants will not be running at their design efficiencies and would rather run at suboptimal, subcritical condition, defeating the purpose of setting up new high efficiency units.

What if Analysis
The Jack-knife test has been used to find out the estimates of the three independent variables ( table 09). Using the estimates given by this method it is found that if peak demand increases by 5000 MW, the PLF will increase by 2.4 % (4.9721e-04 * 5000). If Installed Capacity of coal and lignite increases by 5000 MW, the PLF will decrease by 1.35 % (-2.6945e-04 * 5000) and if total renewable energy capacity increases by 5000 MW, the PLF will decrease by 1.

7.3.2
Rising Costs making thermal power generation un-attractive -The regular inflation and increase in wages of employees and workers is something which is already affecting the O&M cost of thermal power which is manpower intensive industry. Moreover, the major input to thermal power -coal is also becoming costlier.
Following table shows the Wholesale Price Index (Non coking coal) for last seven years. Coal is also facing headwinds due to various other reasons. It is now largely perceived a bad guy all over the world due to the gaseous emission and ash that it produces. More and more pressure to reduce environment footprint is necessitating addition of pollution control equipment. Moreover, Govt has introduced special levy on coal prices.  [34] reported that the main reason coal may have to face battle to fuel India's future energy needs is that it's simply becoming too expensive relative to renewable energy alternatives such as wind and solar. In recent months, power supply auctions have shown that renewables can be offered at less than 3 rupees (4 U.S. cents) per kilowatt hour, a tariff that coal-fired generators have difficulty matching. There is also zero chance that new coal generators can produce electricity at rates competitive to renewables, given higher capital and operating costs.
We argue that this situation is likely to continue and coal is likely to face even more costs thus creating further pressures on PLF. All existing CT Plants must reduce specific water consumption upto 3.5m3/MWh. (iii) All plants installed after 1st Jan'17 must achieve maximum specific water consumption upto 2.5m 3 /MWh and achieve zero waste water discharge.
These are stiff norms are comparable to strict world standards. Further to this, India has made strong commitments to the world body (UNFCCC) through its Nationally Determined Contributions (NDCs) to reduce Global Warming.
India has pledged to improve the emissions intensity of its GDP by 33  With rising concern about pollution, it is felt that the challenges for thermal generators on environment front are likely to remain or be even more stringent. The PLFs projected in this study can be even further negatively affected due to the stringent environmental conditions.  [8] reported that Indian government has raised the minimum quantity of renewable power that states must procure to 21% of their overall power purchases by FY22. It is pertinent to add here that this RPO obligation is terms of energy (kWh) and not in terms of power (MW). The implication is that since the renewable energy is not available round the clock, the purchasing entities (Discoms) have to buy higher percentage in terms of MW whenever renewable energy is available. This could be as high as 50-60 % in MW terms. When such large quantum is bought from renewables, thermal has to reduce generation (ramp down). At other times in a day, when renewable is not available (say in the evening peak time), then thermal power will have to ramp up generation to meet the demand. This situation forces thermal power plants to run a cyclic operation or flexible operation regime where they have to ramp up ramp down generation several time a day. This situation will continue and will become even more stringent. The Economic Times (2019, Oct 06) [39] reported that the Delhi's power regulator DERC has rate of return on equity shall be reduced by 0.25% in case of failure to achieve the ramp rate of 1% per minute; b) an additional rate of return on equity of 0.25% shall be allowed for every incremental ramp rate of 1% per minute achieved over and above the ramp rate of 1% per minute, subject to ceiling of additional rate of return on equity of 1.00%. [40] Faced with this new situation, thermal power plants are bracing themselves for flexible operation. This requires major changes in maintenance strategy and operation strategy. It will also require investment in retrofitting some equipment with additional functionality and monitoring. If such changes are not done, power plant units will either trip on fault or will have to be put under shutdown. This will result in further fall of plant Utilization Factor (PLF). It means that while PLFs will be down due to ramp down situation, it can go drastically down if the plants do not cope up with flexible operation because in that case they will have to be shut down.

P a g e 36 | 46
This is one area where urgent action is required. Thermal  This projection shows that production will reach 871 MMT by 2024-25. The State-run Coal India (The largest and the monopoly coal producer in India) has set itself an annual production target of 1 billion tonnes till 2024 -Ankit Saproo, Economic Times, (2020, Jan 02) [41]. Further, the Govt is planning to auction over 200 coal blocks in the next five years. The move is in accordance with Govt's efforts to lower India's dependence on coal imports. Removing hurdles in commercial coal mining will help in making India self-sufficient in coal production to meet the requirements.
Here we consider the Scenario V ( has several sources to buy power from and also has to meet RPO, it may not buy all the available power from a particular producer. Since the purchaser pays the full fixed charges, it has the right to buy or keep the asset available for buying as the need arises for the purchaser. So the producer is asked to ramp down generation. At the same time, elsewhere in the country, there could be a buyer, who is ready to buy power at the rate which the producer is giving to the contracted buyer or even at a higher price. In absence of a techno commercial mechanism this exchange was not feasible. A new regulation called Real Time Market (RTM) has come a breather in this situation. Under this mechanism, in the event of the power producer not getting full dispatch schedule from contracted buyer, the power producer takes prepermission of the buyer to sell the additional quantum of electricity in open market through trading in electricity stock exchange. The stock exchange connects buyers and sellers through price clearing mechanism in real time in short cycles. The electricity can thus get sold to a willing buyer. Any additional gain made by power producer in the variable cost, will be shared between the power producer and the original contracted buyer (because the original  4. Although the regression model had high R2, the factors considered for regression had high collinearity. Due to this, the Partial Least Square (PLS) method had to be used to predict the future values of PLF.
5. This paper has not studied the emissions arising out of the thermal power station.

Author contribution
This paper has been entirely written by the single author Alok Kumar Tripathi

Conflict of Interest
Annexure 1

Research Questionnaire
As you are aware, utilisation factor [PLF] of coal based thermal power station in the country has a seen a consistent downward trend in last few years and the future outlook is also not very clear. This situation is affecting the power producers, lenders, Discoms, consumers and almost everybody else connected with power sector. This research is an attempt to find out the factors that are responsible for capacity utilisation of coal based plants in India so that they can be addressed and future trend can be predicted.
We seek your expert opinion about rating of factors that are responsible for downward trend/low capacity utilisation of coal based thermal power in India. Please rate the following based on your understanding/ experience.
Please rate the factors as per your opinion whether the factor has -