A power grid is simply one big interconnected machine. Whenever someone turns on a light switch, someone on the other side has to increase production. This is a simple case of supply and demand. However, this balancing act is becoming more uncertain and unpredictable because of the emergence of renewables. So, it requires new solutions such as batteries which can store energy, hydrogen, and even super connected grids linking countries.
However, there is another option – demand response. This refers to changes in the use of electricity by consumers during high periods in order to decrease demand on the power grid and ensure electricity reliability. Therefore, this focuses on process rescheduling instead of new technology investments. It is also more data-driven, which means that AnyLogic simulation could be applied here effectively.
The raw material, which in this case is meat cuttings, is received by the industrial consumer and needs to be stored in a cooled room before it is processed. In the system, there is a maximum and minimum temperature threshold within which the meat can be stored. So, the temperature can be changed depending on the power grid’s needs in order to provide a stable service.
For the electricity supplier, the prices vary depending on the time of the day, for example, the early morning and evening are more expensive, while the consumer has fixed prices. If the consumer accepts flexible prices, they can choose to use more electricity. In this case it means increasing the temperature in the cooling room during the non-peak hours in the grid, and reducing the temperature in the cooling room during peak hours. This will ease the electricity consumption during peak hours.
This analysis of demand response using simulation focused on a single industrial consumer, which was a Danish company in the meat processing industry. With simulation it is possible to compare different scenarios: to vary parameters and see how the modeled system responds.
AnyLogic simulation software was the best option to simulate this demand response effect because of the following advantages it brings to an analysis:
- Descriptive simulation approach in order to compare results of different operational conditions and help decision making in a multivariate program.
- Discrete-event simulation as timing is crucial for market strategy.
- Agent-based modeling, where agent interactions influence information access, which in turn influences market strategy.
Using simulation, the modeler can visualize the results showing the potential of the simulation software in the following ways:
- Evaluate the current potential using a simulation experiment
- The Financial impact of demand response which includes the cumulative and daily results.
- The Process impact of demand response where the cumulative and real-time results are displayed.
- Evaluate different market options using a CompareRuns experiment
- Evaluate operational flexibility using a ParameterVariation Experiment
In this experiment, a market-based explicit direct response strategy was chosen. This means participating in the market directly as a consumer without involving the electricity supplier. It is explicit because it is clear and direct without other assumptions. The consumer is then interested in the results which are broken down into two sections:
Financial impact of demand response - Cumulative results
Process impact of demand response - Cumulative results
Process impact of demand response – Real-time results
Here the user can compare different scenarios, which is where the power of simulation comes in. Users can run various simulations with alternative direct responses, and market configurations. Then see how their electricity bill evolves with time. It is possible to compare the relative savings contrasted to the base rate. Essentially, finding that various market combinations can give different savings.
This example showed an increase in the maximum temperature parameter of half a degree. As a result, the consumer can bid more on the electricity market and make a certain amount of savings. There is, however, a saturation effect. This means that the more the consumer increases the temperature, the less marginal benefits they receive. From this, the industrial consumer would understand that increasing the temperature by another half a degree wouldn’t reduce costs.
The results of this demand response showed that consumers could participate in the electricity market using three approaches:
- Descriptive approach, where the consumer decides what parameters are acceptable to participate in the electricity market.
- Consumer-centric approach, focusing on the results for the industrial consumer and not for the grid operator, yet at the same time continuing to model the logic of all the other market players.
- Modular approach, because different business models are tested, shifting the roles and distributing them to different agents.
Finally, a larger aim has been identified to apply this simulation to a number of different consumers in two different contexts – the Danish and the Chinese market. Here the usefulness of the modular approach can be seen because the two markets are quite different in structure, yet the interfaces can be quickly switched from one agent to the other. These can then be implemented in many different industrial processes.
The case study was presented by Nicolas Fatras, PhD student, Center for Energy Informatics, University of Southern Denmark at the AnyLogic 2021 Conference.
The slides are available as a PDF.
