Abstract:

Electricity, inherently, is a non-storable commodity and has to be consumed at the moment it is produced. Although there are energy storage technologies that can store electricity, none of them is lossless and a certain amount of electricity is lost during charging and discharging. Due to the high penetration of renewable energy sources, which generate variable electricity intermittently, large energy storage is required for the optimal operation of the grid. In recent times, large-scale batteries are making inroads to the power sector but they are still expensive. Using large central electricity storage suffers from disadvantages of capital, operational and maintenance cost and grid degradation due to storage failure. A large, distributed energy storage consisting of a large number of small-scale energy storage systems (ESSs) is advantageous for both the system operators and the consumers.

A survey estimated that electricity consumers in Pakistan had installed over 2.8 million battery-based uninterruptible power supplies (UPSs) in the year 2012-13 with an increasing rate of 4.5% per year. Consumers installed these UPSs due to load shedding problems in Pakistan caused by electricity shortfalls [1]. In recent years, the Government of Pakistan has added generation capacity at a rate of 15.5% per year, almost twice the rate at which demand increased [2]. The study [2] estimated that a large volume of installed capacity will not be utilized in the coming years even at peak demand. The large number of batteries need to be utilized which would, otherwise, will be unusable due to surplus generation capacity. We can create a large energy storage by aggregating a large number of small-scale batteries that are already installed by the consumers.

In this thesis, we propose a model for creating a large, distributed energy storage consisting of many small-scale consumer’s batteries. The model is based on central control where a central controller controls the charging and discharging of the individual batteries. An important part of the model is scheduling a net charge or discharge energy among the individual batteries participating in the distributed storage. Therefore, we also propose a weighted batteries scheduling (WBS) scheme for scheduling energy among the batteries. The WBS scheme assigns a numeric weight to each of the batteries and the energy is scheduled according to the weight of each of the batteries. We evaluated performance with respect to different weights such as the state-of-charge (SOC) of the batteries, batteries energy capacities and priority-based weights. Priority-based weights, which are calculated by prioritizing the batteries with respect to their SOCs, appear to be the most efficient. Priority-based weights schedule energy among the batteries so that it maximally keeps SOCs of all the batteries at the same level thus minimizing the effect on the batteries’ cycle lifetime.

Energy arbitrage is a mechanism of increasing retailer’s profit by exploiting variations in electricity prices. Retailer buys and stores electricity at the time with low wholesale electricity price and sells stored electricity to the end-users during times of high retail price. Due to internal losses of ESS, efficient charging and discharging schedule should be obtained that minimizes the cost of buying electricity and maximizes the retailer’s profit. We used our distributed storage model to calculate arbitrage profit using wholesale electricity price data. Wholesale electricity prices are usually not available in advance. Therefore, we developed a scheme for forecasting electricity prices using Gaussian process regression to study the effect of forecasting on energy arbitrage profit. In addition to energy arbitrage, we also calculate financial benefits obtained by the proposed distributed energy storage from the frequency regulation market. We perform an economic analysis from the perspective of energy storage owners by calculating the net-present-value for each of the storage owners. Our results show that the proposed model gives financial benefits to both the retailers as well as the consumers. We compare financial benefits obtained by the storage owners using the storage individually i.e. without participating in the distributed storage with the financial incentives obtained by participating in the distributed storage. In the individual case, the ESS remains highly underutilized, and the financial benefits are much lower compared to the financial benefits that a storage owner receives by participating in the distributed storage. We used the proposed storage model for peak shaving applications as well. Using Pakistan demand data, we show that the operational cost can be greatly reduced by shifting a peak from peak hours to off-peak hours using our proposed distributed storage.

Publications:

1. Mehmood, Nasir, and Naveed Arshad. "Towards Developing a Large Distributed Energy Storage Using a Weighted Batteries Scheduling Scheme." IEEE Access 8 (2020): 210733-210749.

2. Mehmood, Nasir, and Naveed Arshad. "Interval Forecasting of Hourly Electricity Spot Prices using Rolling Window Based Gaussian Process Regression." 2020 2nd International Conference on Smart Power & Internet Energy Systems (SPIES). IEEE, 2020.

3. Mehmood, Nasir, and Naveed Arshad. "Economic analysis of using distributed energy storage for frequency regulation." Proceedings of the Eleventh ACM International Conference on Future Energy Systems. 2020.

4. Mehmood, Nasir, and Naveed Arshad. "Towards developing a large energy store using small scale distributed batteries." Proceedings of the 5th Conference on Systems for Built Environments. 2018.