Analyzing electricity smart grids and markets with simulation

Analyzing electricity smart grids and markets with simulation

Like many around the world, the Dutch energy system dates from an era of centralized markets and pre-renewable energy generation. This situation is now changing rapidly – energy markets are open to competition, households commonly generate electricity from renewables, and there is a growing shift to electric vehicles. While these changes take place, electricity delivery must remain stable and affordable.

Sjors Hijgenaar proposes a response to these developments as radical as the changes themselves – decentralized smart grids and direct local energy trading. In his analysis, an agent-based simulation of a Dutch neighborhood demonstrates the benefits in efficiency and reliability of the proposal. How can this work?

Good neighbors

The changing nature of energy generation and consumption, in connection with renewables and electric vehicles, is causing the Netherland’s traditional day-ahead method of purchasing electric to come under strain. Difficulties forecasting demand and setting production mean shortfalls must be purchased on the intra-day market at an elevated cost.

The agent-based simulation used in the analysis replicates market demands by modeling numerous individual agents and subjecting them to business logic. Set in a neighborhood where electric vehicle charging and renewable energy generation are common, the model contains household agents, public charging stations, and household-owned PEV (plug-in electric vehicles).

The simulation shows that while electric vehicles place additional demands on the electric grid, they also provide a resource that can store and deliver energy on demand. Storing and providing electricity for the grid, when needed, increases stability and helps balance demand. There are, however, other considerations, as you may expect.

Apart from meeting the demands of the electricity grid, using the batteries of parked up electric vehicles raises many questions. Not least, those regarding battery degradation, energy pricing, and whether vehicles are ready to drive when needed.

The solution comes from local trading between members of the neighborhood grid and the algorithm to manage it. The proposal suggests direct peer-to-peer trading can be quicker and more efficient than relying on a third party to manage trades. And, if the benefits are clear, neighborhood households will be more likely to take part. To investigate this, the simulation model was connected to a Tendermint blockchain network and resulted in a proof-of-concept local energy market.

A first step?

With the right implementation, local energy trading can help reduce the deviations between estimated and actual consumption that occur in today’s centrally traded energy markets.

Furthermore, the algorithm developed from the analysis successfully considers many other parameters, including minimum state-of-charge for PEV departure, and keeping charging session costs constant under fixed market tariffs. It is a first of its kind algorithm that both incorporates large numbers of individual user preferences as well as benefiting the whole system through peak shaving and shifting.

The proof-of-concept offers grid operators a first step towards safe local energy trade.

Take a deeper dive

You can find much more detail, including the simulation modeling and blockchain market setup, in the full thesis paper Electric vehicles; the driving power for energy transition.

Would you offer your car battery for use on the grid? Would you prefer a decentralized peer-to-peer approach or a centrally brokered system? Have your say in the comments below!

Related posts