1. The energy market of India is rapidly transforming with large scale electrification schemes ensuring 24x7 power for all, increasing penetration of RE and tighter policy measures to ensure smooth operation of national grid.

2. Owing to such reforms the power purchase cost of Utilities has significantly increased over the last few years. This is driving the requirement for utilities to predict the electricity demand precisely and implement optimized generation schedules so that the predicted demand is met at least cost taking into account all available power resources like long term PPA, TAM, DAM and RTM.

3. However accurately predicting 15 minute demand for any given time horizon, forecasting prices in the TAM/ DAM and RTM, selecting the plants for implementing least cost generation schedule and manually preparing bids and managing trading in various electricity markets can be manually intensive for operators relying on inaccurate weather data, spreadsheet based tools and file based approval processes. For example managing 24x7 trading in the real time market with 48 bidding sessions with all necessary approvals for every buy and sell decision daily can be a resource exhaustive task and can lead to errors in the bidding process by operators.

4. This is where the Mercados solution JouleOS comes in. JouleOS provides a comprehensive cloud based machine learning driven power portfolio management operating system that allows the utilities to forecast demand and schedule power in order to fulfill the demand at least cost.
(A) JouleOS Demand Forecast:

Mercados propriety machine learning Forecast engine uses historical demand data, satellite based Global High-Resolution Atmospheric Forecasting data (temperature, wind speed, wind gust, rainfall, humidity), local station-based weather data and data of special days to calculate the 15 minute time block wise demand forecast on intra-day, day-ahead and week-ahead basis.

5. JouleOS Price Forecast:

While the demand forecast engine is calculating the expected day ahead demand, the Mercados deep learning price forecast engine automatically collects market data to estimate the market clearing price in the day ahead market on the Power Exchange.

  • 6.(A) Once the day ahead demand and price forecast is available JouleOS dispatch optimization engine gathers
  •       1. inter-state and intra-state generation availability through a cloud based API and
  •       2. purchase and sell opportunities on the power exchanges based on the JouleOS price forecast of power exchange and
  • (B) Establishes a time-block wise schedule to fulfill the predicted demand

7. The buy and sell decision is finalized through a cloud based cross platform approval process so that every buy or sell decision is ratified by the power management team of the utility.

8. Once approved the JouleOS trading engine places bids on the Power Exchanges through automated cloud based API.

9. Once the bid results are declared the JouleOS automatically sends the final generation schedule to the grid operators for implementation well in time before gate closure.

10.(A) The entire process can be implemented for both day ahead and real time market for as many bidding sessions as the regulations may stipulate.

(B) In case of any outages or change in forecasted demand the JouleOS despatch engine automatically detects the changes and re-runs the optimization engine to modify the generation schedule taking into account RTM prices and sends revised implementation schedule to the grid operators.

11.(A) JouleOS provides a comprehensive dashboard to all the users that gives them complete control over the real time market dynamics and allows near to gate closure planning and despatch.

(B) JouleOS is supported by a 24x7 analytics team of Mercados.

(C) JouleOS is complaint with all regulatory standards defined in various regulations includes IEGC and Open access regulations.

(D) Utilities that deployed JouleOS have already witnessed significant reduction in power purchase cost and efficient planning of resources.