US-12626268-B2 - Method and system for optimizing operation and price of an energy storage as a service (ESaaS)

US12626268B2US 12626268 B2US12626268 B2US 12626268B2US-12626268-B2

Abstract

The embodiments of present disclosure address a need of a framework to holistically utilize storage capacity of an Energy Storage System (ESS) to serve forecast errors of several Renewable Energy Generators (REGens) participating in a day-ahead market. Embodiments herein provide a method and system for optimizing the operation and price of an Energy Storage as a Service (ESaaS) framework. In anticipation of the forecast errors from REGens, the ESS operator takes suitable countermeasures such as charging/discharging of storage system through market transactions. This is done in a way to reduce imbalance in the market commitments made by individual REGens without reserving any storage volume for each REGen. Further, the system is configured to schedule the storage, determine the settlement volumes, and decide the service prices. The disclosed ESaaS framework is beneficial for all entities such as REGens (revenue outflow decreases), system operator (imbalance volume reduces), and ESS (revenue earned increases).

Inventors

Vishnu Padmakumar Menon
Yogesh Kumar BICHPURIYA
Narayanan RAJAGOPAL
Venkatesh Sarangan

Assignees

TATA CONSULTANCY SERVICES LIMITED

Dates

Publication Date: 20260512
Application Date: 20240507
Priority Date: 20230516

Claims (9)

1 . A processor-implemented method comprising: receiving, via input/output interface, a historical error data of one or more Renewable Energy Generators (REGens), wherein the historical error data is a deviation between an actual generation and a committed generation; aggregating, via one or more hardware processors, the received historical error data of the one or more REGens, wherein a REGen i, interested in a best-effort energy storage service, shares historical values of its market deviation error signal with an Energy Storage System ESS and the ESS determines an appropriate service price, wherein based on the historical values and real-time updates of the market deviation error signal, the ESS offers a best-effort Energy Stored as a Service ESaaS for REGen i; training, via the one or more hardware processors, a Long Short-Term Memory (LSTM) network using the aggregated historical error data to forecast an error of a required day, wherein when the aggregate historical error estimates are positive the ESS absorbs a predefined quantum of energy from a grid to charge the storage of the ESS, wherein when the aggregate historical error estimates are negative the ESS compensates by supplying a predefined quantum of energy to the grid by discharging the storage of the ESS; training, via the one or more hardware processors, a Hidden Markov Model (HMM) on the aggregated historical error data to generate one or more error samples and to obtain a representative error profile using a distance metric among the error samples; assigning, via the one or more hardware processors, a performance score to each of the one or more REGens based on the one or more statistical error properties of each of the one or more REGens; determining, via the one or more hardware processors, a service price of each of the one or more REGens based on the received historical error data for maximizing revenue of an Energy Storage System (ESS) and acceptance likelihood (pi) of the one or more REGens, wherein the determined service price is weighted with the assigned performance score of each of the one or more REGens to get a final price per unit of the forecasted error served, and wherein the accepted likelihood (pi) of the one or more REGens is determined to accept the service price via a sigmoid function, wherein when the REGen i accepts the price offer, the REGen i starts sharing the updated error forecasts for each time slot obtained just before the delivery with the ESS B; determining, via the one or more hardware processors, a schedule of charging and discharging of storage of the ESS and market commitments of the ESS in a day-ahead market based on the representative error profile and determined service price; obtaining, via the one or more hardware processors, an actual deviation from the one of more REGens in real time; modifying, via the one or more hardware processors, the schedule of charging and discharging of storage of the ESS based on the obtained actual deviation of the one or more REGens, associated market commitments of the ESS to buy and sell in the day-ahead market and the determined service price; and determining, via the one or more hardware processors, actual served errors and unserved errors of each of the one or more REGens by the ESS and a deviation created by the ESS based on modified schedule of charging and discharging and the market commitments of the ESS in the day-ahead market, wherein an ESS operator dynamically charges or discharges storage in anticipation of the deviations in the generation volume of its subscribers and the storage charge or discharge schedule is adjusted to account for inaccuracies, wherein the ESS operator dynamically leverages the unused storage volume for other applications.
2 . The processor-implemented method of claim 1 , wherein the ESS traded-off aggregated market deviation error in a day-ahead market to reduce imbalance.
3 . The processor-implemented method of claim 1 , wherein one or more statistical error properties comprises a time average, a time deviation and a maximum temporal correlation.
4 . A system comprising: an input/output interface to receive a historical error data of one or more Renewable Energy Generators (REGens), wherein the historical error data is a deviation between an actual generation and a committed generation; at least one memory in communication with the one or more hardware processors, wherein the one or more hardware processors are configured to execute programmed instructions stored in the at least one memory to: aggregate the received historical error data of the one or more REGens, wherein a REGen i, interested in a best-effort energy storage service, shares historical values of its market deviation error signal with an Energy Storage System ESS and the ESS determines an appropriate service price, wherein based on the historical values and real-time updates of the market deviation error signal, the ESS offers a best-effort Energy Stored as a Service ESaaS for REGen i; train a Long Short-Term Memory (LSTM) network using the aggregated historical error data to forecast an error of a required day, wherein when the aggregate historical error estimates are positive the ESS absorbs a predefined quantum of energy from a grid to charge the storage of the ESS, wherein when the aggregate historical error estimates are negative the ESS compensates by supplying a predefined quantum of energy to the grid by discharging the storage of the ESS; train a Hidden Markov Model (HMM) on the aggregated historical error data to generate one or more error samples and to obtain a representative error profile using a distance metric among the one or more error samples; assign a performance score to each of the one or more REGens based on the one or more statistical error properties of each of the one or more REGens; determine a service price of each of the one or more REGens based on the received historical error data for maximizing revenue of an Energy Storage System (ESS) and acceptance likelihood (pi) of the one or more REGens, wherein the determined service price is weighted with the assigned performance score of each of the one or more REGens to get a final price per unit of the forecasted error served, and wherein the accepted likelihood (pi) of the one or more REGens is determined to accept the service price via a sigmoid function, wherein when the REGen i accepts the price offer, the REGen i starts sharing the updated error forecasts for each time slot obtained just before the delivery with the ESS B; determine a schedule of charging and discharging of storage of the ESS and market commitments of the ESS in a day-ahead market based on the representative error profile and determined service price; obtain an actual deviation from the one of more REGens in real time; modify the schedule of charging and discharging of storage of the ESS based on the obtained actual deviation of the one or more REGens, associated market commitments of the ESS to buy and sell in the day-ahead market and the determined service price; and determine actual served errors and unserved errors of each of the one or more REGens by the ESS and a deviation created by the ESS based on modified schedule of charging and discharging and the market commitments of the ESS in the day-ahead market, wherein an ESS operator dynamically charges or discharges storage in anticipation of the deviations in the generation volume of its subscribers and the storage charge or discharge schedule is adjusted to account for inaccuracies, wherein the ESS operator dynamically leverages the unused storage volume for other applications.
5 . The system of claim 4 , wherein the ESS traded-off aggregated market deviation error in a day-ahead market to reduce imbalance.
6 . The system of claim 4 , wherein one or more statistical error properties comprises a time average, a time deviation and a maximum temporal correlation.
7 . One or more non-transitory machine-readable information storage mediums comprising one or more instructions which when executed by one or more hardware processors cause: receiving, via input/output interface, a historical error data of one or more Renewable Energy Generators (REGens), wherein the historical error data is a deviation between an actual generation and a committed generation; aggregating, via one or more hardware processors, the received historical error data of the one or more REGens, wherein a REGen i, interested in a best-effort energy storage service, shares historical values of its market deviation error signal with an Energy Storage System ESS and the ESS determines an appropriate service price, wherein based on the historical values and real-time updates of the market deviation error signal, the ESS offers a best-effort Energy Stored as a Service ESaaS for REGen i: training, via the one or more hardware processors, a Long Short-Term Memory (LSTM) network using the aggregated historical error data to forecast an error of a required day, wherein when the aggregate historical error estimates are positive the ESS absorbs a predefined quantum of energy from a grid to charge the storage of the ESS, wherein when the aggregate historical error estimates are negative the ESS compensates by supplying a predefined quantum of energy to the grid by discharging the storage of the ESS; training, via the one or more hardware processors, a Hidden Markov Model (HMM) on the aggregated historical error data to generate one or more error samples and to obtain a representative error profile using a distance metric among the error samples; assigning, via the one or more hardware processors, a performance score to each of the one or more REGens based on the one or more statistical error properties of each of the one or more REGens; determining, via the one or more hardware processors, a service price of each of the one or more REGens based on the received historical error data for maximizing revenue of an Energy Storage System (ESS) and acceptance likelihood (pi) of the one or more REGens, wherein the determined service price is weighted with the assigned performance score of each of the one or more REGens to get a final price per unit of the forecasted error served, and wherein the accepted likelihood (pi) of the one or more REGens is determined to accept the service price via a sigmoid function, wherein when the REGen i accepts the price offer, the REGen 1 starts sharing the updated error forecasts for each time slot obtained just before the delivery with the ESS B; determining, via the one or more hardware processors, a schedule of charging and discharging of storage of the ESS and market commitments of the ESS in a day-ahead market based on the representative error profile and determined service price; obtaining, via the one or more hardware processors, an actual deviation from the one of more REGens in real time; modifying, via the one or more hardware processors, the schedule of charging and discharging of storage of the ESS based on the obtained actual deviation of the one or more REGens, associated market commitments of the ESS to buy and sell in the day-ahead market and the determined service price; and determining, via the one or more hardware processors, actual served errors and unserved errors of each of the one or more REGens by the ESS and a deviation created by the ESS based on modified schedule of charging and discharging and the market commitments of the ESS in the day-ahead market, wherein an ESS operator dynamically charges or discharges storage in anticipation of the deviations in the generation volume of its subscribers and the storage charge or discharge schedule is adjusted to account for inaccuracies, wherein the ESS operator dynamically leverages the unused storage volume for other applications.
8 . The one or more non-transitory machine-readable information storage mediums of claim 7 , wherein the ESS traded-off aggregated market deviation error in a day-ahead market to reduce imbalance.
9 . The one or more non-transitory machine-readable information storage mediums of claim 7 , wherein one or more statistical error properties comprises a time average, a time deviation and a maximum temporal correlation.

Description

PRIORITY CLAIM This U.S. patent application claims priority under 35 U.S.C. § 119 to Indian application Number 202321034354, filed on May 16, 2023. The entire content of the abovementioned application is incorporated herein by reference. TECHNICAL FIELD The disclosure herein generally relates to the field of energy storage services (ESS) and more specifically, to a method and system for optimizing operation and price of an energy storage as a service (ESaaS). BACKGROUND Transiting to a sustainable economy mandates tight integration of renewable energy generation with mainstream power grids. These Renewable Energy Generators (REGens) earn revenue by selling their power output to others. Since bilateral contracts are less lucrative in the short-term, the REGens focus on electricity markets also to auction their power. As trading in intraday markets requires a more sophisticated set-up, the REGens target day-ahead markets too. However, nature induced stochastic variations in generation introduces risks in terms of volume commitments that can be made by the REGens in day-ahead markets. Any deviation from the commitments in the day-ahead market leads to penalties or settlement at imbalance prices. The REGens can minimize the risk of deviating from market commitments with the help of an energy storage system (ESS). The ESS storage can be charged to store excess power during over-generation and discharged during periods of under-generation. However, most of existing set-ups consider ESS either dedicated for one REGen or shared across a group of REGens. When the ESS is shared across REGens, almost all the works reserve a fraction of storage volume for each REGen. SUMMARY Embodiments of the disclosure present technological improvements as solutions to one or more of the above-mentioned technical problems recognized by the inventors in conventional systems. For example, in one embodiment, a method and system for optimizing price and operation of an energy storage as a service (ESaaS) is provided. In one aspect, a processor-implemented method for optimizing price and operation of an energy storage as a service (ESaaS) is provided. The processor-implemented method includes one or more steps such as receiving, via input/output interface, a historical error data of one or more Renewable Energy Generators (REGens), wherein the historical error data is a deviation between an actual generation and a committed generation. Further, the processor-implemented method includes aggregating the received historical error data of the one or more REGens, training a Long Short-Term Memory (LSTM) network using the aggregated historical error data to forecast an error of a required day and training a Hidden Markov Model (HMM) on the aggregated historical error data to generate one or more error samples to obtain a representative error profile using a minimization of least-absolute distance among the error samples. Furthermore, the processor-implemented method comprising assigning a performance score to each of the one or more REGens based on one or more statistical error properties of each of the one or more REGens, determining a service price of each of the one or more REGens based on the received historical error data using an optimization framework for maximizing revenue of the ESS and the one or more REGens acceptance likelihood and determining, via the one or more hardware processors, a schedule of charging and discharging of storage of the ESS and market commitments of the ESS in a day-ahead market based on the representative error profile and determined service price. Further, the processor-implemented method comprising obtaining an actual deviation from one of more REGens in real time and modifying the schedule of charging and discharging of storage of the ESS based on the obtained actual deviation of the one or more REGens, associated market commitments of the ESS of buy and sell in the day-ahead market and the determined service price. Finally, the actual served errors and unserved errors of each REGens by ESS and deviations created by ESS are determined based on modified schedules and ESS market commitments in the day-ahead market. In another aspect, a system for optimizing price and operation of an energy storage as a service (ESaaS) is provided. The system includes an input/output interface configured to receive a historical error data of one or more Renewable Energy Generators (REGens), wherein the historical error data is a deviation between an actual generation and a committed generation, one or more hardware processors and at least one memory storing a plurality of instructions, wherein the one or more hardware processors are configured to execute the plurality of instructions stored in the at least one memory. Further, the system is configured to aggregate the received historical error data of the one or more REGens, train a Long Short-Term Memory (LSTM) network using the aggregated historical error data to forecast an