The StRe@M model is able to capture and determine the stackable revenues of flexibility resources
The different sources of flexibility may have technological and cost-related barriers to prevent their uptake. There are also barriers in regulation, standardization and energy market rules. However, regulations, standards and markets are being developed worldwide to facilitate emerging flexibility solutions, like energy storage and demand response. One of the standardization initiatives is DNV GL’s recommended practice on energy storage, GRIDSTOR. For most flexibility resources, especially energy storage, a single flexibility service will not ensure a positive business case.
Stackable revenues, by combining operational services in different markets and time scales, are therefore important to render the business case worthwhile. DNV GL has developed the StRe@M model, which is able to assess and analyse the short and long term business case of flexibility resources whilst taking into account stackable revenues. The model is able to capture and determine the stackable revenues of flexibility resources such as a Li-ion battery, demand response, a gas engine or a pumped hydro-electric storage plant. The StRe@M model optimizes the portfolio of flexibility services, while considering uncertainty in the forecasts of the market prices and RES generation for the next hour(s), day(s), and longer periods.
A business case tool for flexibility resources
A cost-benefit analysis is of key importance when assessing flexibility options. This is especially true for storage systems, because they generally compete with other flexibility solutions. Moreover, multiple services should be considered for storage, to optimize its business case (or even make it positive). Quantifying revenues and defining operational strategies over various markets and services presents a complex problem, not in the least because of the uncertainty in future market and pricing developments. For analysis of these business cases, versatile tools are needed. One such tools is the StRe@M model developed by DNV GL. This tool can assess the short and long-term business case of flexibility resources while considering stackable revenues. StRe@M is presented in this chapter where several study cases for the business case analysis are described.
There are five modules in the STREAM approach:
- Spot market forecast
First, a (perfect) forecast of the spot market prices is obtained. For this analyses, historical APX spot prices of 2015 have been used
- Imbalance forecaster
The spot price time series is used in the ‘Imbalance Forecaster’ module: this module generates multiple synthetic imbalance volume and balancing price time series based on 2015 balancing market data from the Dutch TSO (TenneT). These synthetic imbalance volumes and prices represent the uncertainty (and opportunities) on the balancing market at the time of spot market trade optimization
- Spot market bid optimization
This module optimizes the spot market trades of the battery, given the (perfect) forecast of the spot market prices and the (imperfect) forecasts of the balancing market. Also, the optimization takes the relevant techno-economical parameters of the battery into account, such as maximum charge and discharge capacity, state of charge, and the impact of the depth of discharge on cycle lifetime. The output of the optimization is a single time series for selling and purchasing electricity from the day-ahead spot market. The optimization is performed in steps of one day with a time-resolution of 15-minutes
- Real-time dispatch
The real-time dispatch optimizes the real-time dispatch of the battery per PTU given a (perfect) forecast of the balancing prices and volumes for the next hour (i.e. only short-term foresight). However, the obligations with respect to the spot market are well known in advance and are implemented as a constrained that must be met. The real-time dispatch is performed separately for each balancing time series from the Imbalance forecaster module
- Investment analyses
The investment module performs a Monte Carlo analyses based on the uncertainty in the investment costs and the uncertainty in the combined revenues from the spot and balancing market. As the revenues are determined for one specific year, the investment costs are annuitized and an annuitized net-present-value probability distribution is calculated.