Comments on the Proposed Methodology for Capacity Credit of Generation Resources and Coincident Peak Requirements

Having said that, there are some issues, if addressed, will make the methodology fully acceptable to all stakeholders including small states. I have detailed the pros and cons of the proposed methodology for consideration:

Pros:

1. Informed Decision-Making: The methodology provides detailed frameworks, such as the Top 5% demand hours and coincident peak methods, which enable state utilities to accurately assess and plan for resource adequacy. This approach allows for strategic energy planning to ensure grid reliability.

2. Handling VRE Variability: It introduces scientific approaches to calculate the capacity credit of Variable Renewable Energy (VRE) sources, addressing the intermittency challenge of renewables like solar and wind. The methodology acknowledges the need for time-dependent assessment and seasonal analysis to accommodate fluctuating energy generation patterns.

3. Flexibility: The methodology is adaptable for both solar and non-solar hours, recognizing the differing generation profiles across these periods. This is critical in ensuring that utilities account for energy generation variations across different times of the day and seasons.

4. Capacity Credit Evaluation: By introducing the Effective Load Carrying Capability (ELCC) and other methods, the methodology provides a robust framework for evaluating the true dependable output of renewable sources, reducing over-reliance on nameplate capacity which can be misleading for VREs.

5. Avoiding Overcapacity Risks: The methodology avoids the pitfalls of overcapacity by focusing on coincident peak analysis, ensuring that utilities do not overestimate their needs based solely on their own peak demands. This reduces the risk of investing in underutilized infrastructure, which could lead to increased costs.

6. Global Alignment: The methodology aligns with best practices from other countries that have already integrated high levels of renewable energy, making it easier for Indian utilities to adopt tried and tested techniques in resource adequacy planning.

7. Reduction in Overestimation Risks: By employing percentile-based aggregation (80th percentile) instead of maximum values, the methodology reduces the risk of overestimating the required capacity at the national level. This percentile approach ensures that extreme outlier demand values from individual states do not disproportionately inflate the national peak estimate.

8. Improved Precision for Diverse Grids: The differentiation between solar and non-solar hours helps to mitigate aggregation errors specific to renewable energy sources. By isolating periods of variability, such as non-solar hours, the methodology provides a more realistic assessment of renewable energy contributions, ensuring that national-level planning remains grounded in actual operational conditions.

9. Coincident Peak Methodology: The focus on coincident peaks across states reduces aggregation errors, as it aligns state contributions to the national peak. This avoids overestimating generation needs by synchronizing state demand peaks, which otherwise might differ due to localized consumption patterns.

10. Historical Data Usage: The use of historical demand profiles over several years provides a more stable and robust foundation for aggregation. This minimizes the likelihood of short-term anomalies skewing national-level estimations, thus making the approach more reliable for long-term planning.

Cons:

1. Data-Intensive: The proposed methodology, especially approaches like the Top 5% demand hours and Critical Day Analysis, relies heavily on high-resolution data and advanced computational resources. Smaller utilities or those with limited technical capacity might struggle to implement these complex calculations effectively. They need to be provided with technical assistance.

2. Increased Complexity: The methodology, while comprehensive, adds layers of complexity by requiring separate analysis for solar and non-solar hours, especially when integrating VRE sources. This may overwhelm stakeholders or result in miscalculations, especially for states with less technical expertise in energy forecasting.

3. Challenges with Non-Solar Hours: The methodology’s treatment of non-solar hours in demand analysis may disadvantage states with higher solar capacity, as the non-solar hour capacity credit could be significantly lower, leading to challenges in meeting coincident peak demand during these hours especially winter nights when VRE is minimal.

4. Over-reliance on Historical Data: Capacity credit estimation, particularly for conventional sources, depends heavily on historical generation data, which might not accurately reflect future performance due to changes in fuel availability, grid conditions, or policy shifts life enhanced share of RPO/Storage obligation etc. This could lead to inaccurate capacity credit assignments for some resources.

5. VRE Specific Issues: For VRE sources, factors like location specificity, DC/AC ratios, and seasonal variability introduce uncertainties into capacity credit estimations. The methodology recognizes these challenges but does not provide foolproof solutions to mitigate their impact.

6. Potential for Overcompensation: While the methodology prevents overcapacity risks, it might also lead to under-compensation for renewables, especially when capacity credit factors are conservatively assigned. This could slow down the renewable energy transition by disincentivizing VRE investments.

7. Potential Underestimation in Non-Coincident Peaks: By focusing on coincident peak demand hours, the methodology might underestimate the overall capacity needed during non-coincident peak periods, especially in states with significant seasonal or regional demand variations. This can lead to aggregation errors that could impact grid reliability during non-peak but still high-demand periods.

8. Uniform Capacity Credit Application: The capacity credit assignment at a national level could result in aggregation errors if the variability in state-level renewable generation (e.g., wind and solar) is not adequately captured. Since renewable generation varies significantly by geography and season, applying a uniform capacity credit could lead to overestimation or underestimation of actual firm power available from renewables.

9.  Over-Simplification with Averaging: While percentile methods reduce extreme cases, they can still mask regional or state-specific spikes in demand. Averaging coincident demand values over a set period may fail to account for short, sharp peaks in certain states, leading to aggregation errors that result in either surplus capacity or shortfalls during critical periods. An example of over-simplification with averaging could be seen in a situation where Uttar Pradesh (UP) and Tamil Nadu have significantly different electricity demand patterns. Suppose UP experiences a sudden, sharp peak in demand due to extreme weather conditions (e.g., a heatwave), causing its peak demand to spike sharply during the summer months. Tamil Nadu, on the other hand, might have a more consistent demand pattern with smaller fluctuations throughout the year.

If the national demand profile is averaged across states using a percentile method, this sharp, localized peak in UP could be smoothed out when aggregated with Tamil Nadu’s steadier demand profile. The result is that the national coincident peak calculation might not fully capture UP’s extreme peak. This leads to an underestimation of the capacity needed to meet UP’s actual peak demand, risking shortfalls during critical periods when UP’s grid is stressed. Conversely, if Tamil Nadu’s lower demand during these periods is averaged in, the methodology might suggest surplus capacity in Tamil Nadu, where it isn’t needed, potentially leading to inefficient resource allocation and excess capacity in that region.

Let’s assume two states, Uttar Pradesh (UP) and Tamil Nadu (TN), have different peak demand patterns over the course of a week. Below is the hypothetical peak demand (in MW) for each state over seven days:

Step 1: Averaging Across Days

If we use a simple average across these days to calculate the national coincident demand, the results would be:

– UP Average Demand = (22,000 + 23,000 + 35,000 + 24,000 + 23,500 + 22,500 + 23,000) / 7 = 24,143 MW

– TN Average Demand = (15,000 + 14,500 + 15,500 + 16,000 + 15,000 + 16,500 + 15,500) / 7 = 15,429 MW

The national coincident peak is the sum of the averages:

– Total Average Demand (National) = 24,143 MW + 15,429 MW = 39,572 MW

Step 2: Ignoring the Sharp Peak

Now, consider Day 3, where UP experiences an extreme demand spike of 35,000 MW due to a heatwave. The averaging method reduces the significance of this sharp peak because it smooths the demand over the whole week.

If we look at Day 3 alone:

– UP Demand = 35,000 MW

– TN Demand = 15,500 MW

– Total Peak Demand (National) on Day 3 = 35,000 MW + 15,500 MW = 50,500 MW

On this particular day, the national grid needs to supply 50,500 MW. However, using the average method across the week, the coincident demand estimate is 39,572 MW. The shortfall during the real peak is:

– Shortfall = 50,500 MW – 39,572 MW = 10,928 MW

Impact of Over-Simplification

Due to the averaging approach, the methodology suggests that 39,572 MW is the required capacity at the national level, which does not account for the actual peak demand of 50,500 MW on Day 3. As a result, there is a shortfall of 10,928 MW during the heatwave. If the system is planned based on the average, there would be insufficient capacity to meet the peak demand during such extreme periods, leading to potential grid stress or power outages in UP.

Above is an example of how oversimplifying with averages can lead to aggregation errors, underestimating actual peak requirements in certain states and risking grid reliability and how regional spikes can be masked by averaging, potentially creating misalignments in the national-level capacity planning.

10. Limited Consideration of Localized Factors: Although historical profiles are used, the methodology may not fully account for future regional developments such as industrial growth, urbanization, or energy policy changes that could significantly affect demand at the state level. These localized factors might introduce aggregation errors when combining state-level data into national estimates.

Conclusion:

The proposed methodology for assessing the capacity credit of generation resources and determining coincident peak requirements is robust and scientifically grounded, particularly in addressing the variability of VREs and makes reasonable efforts to reduce aggregation errors. However, it introduces significant complexity and demands substantial technical resources, which might pose challenges for utilities with limited expertise and there remain potential issues with under- or overestimation due to reliance on averaging, coincident peak focus, and uniform capacity crediting across diverse state energy profiles. Balancing accuracy with simplicity should be a key consideration in the further refinement of this methodology.