By smartgrider In DEMAND SIDE MANAGEMENT Posted 2015-05-24  (No Comments)


Project Ownership ● DNV KEMA - Energy consulting and testing &certification company is responsiblefor the overall project management and design, expansion of the demonstration project and cost-benefit validation.
● Enexis - Distribution network operator explores the feasibility and economic benefits of capacity management in the distribution network.
● Essent - Energy supplier focuses on smart energy services for the end-users, including the market processes and economics plus smart charging of the electricvehicles [EVs].
● Gasunie - Gas infrastructure company that contributes to research towards an affordable sustainable energy supply, in which gas plays a role in keeping the system as a whole efficient and affordable.
● ICT Automatisering - Software company is responsible for the design, realization,management and maintenance of the ICT infrastructure.
● TNO - Knowledge institute develops and provides the PowerMatcher technology
● TU/e [Electrical faculty] - The faculty is providing the models for capacity management and the integrity of the electricity supply.
● TU Delft [Industrial design faculty] - This faculty is arranging the end-user research.
● Hanzehogeschool - is contributing to the end-user research.
Focus of the projectA new smart coordination mechanism that moves the energy consumer to the center of the energy market.
Physical elements of the project142 TWh
Electricity supplied- 2012● 40 households with solar pv, heat pumps/micro-chp with storage, ‘smart’ devices and HEM
● 10 electrical vehicles
● 2 smart distribution transformers
New servicesNew tariff structures giving different kinds of added value to customers.
Electricity system in the NetherlandsOutage: < 35 minutes/year [housholds]
Consumption households: 3.500 kWh/year
Connection on household level to both electricity and natural gas grid: 96%
Consumption total in 2011: 120 kWh
Production gas fired generation 2010: 63%
Production coal fired generation 2010: 22%
Production wind and solar 2011: 4%
Target 2020 renewables: 14%
Expected % renewable electricity 2020: > 30%
Market structureThe Netherlands [16, 7 million people] have a liberalized market with independent network operators. One TSO, 8 DSO’s.
ContactErik ten Elshof Mail: E.J.tenElshof@minez.nl

The Netherlands


This project was completed successfully and is now getting a follow-on stimulus by the Dutch Government: PowerMatching City II, with an additional 18 households, 10 electric vehicles (EVs) and 2 smart distribution transformers. Hoogkerk is thereby gaining practical experience with new tariff structures and the feeding in of renewable energy into the network, amongst other things.

PowerMatchting City II is one of the twelve trials in the Netherlands aimed at accelerating the Smart Grid development. The main focus of these trials is gaining real world experiences with different Smart Grid applications and real users. New technologies, new partnerships and new forms of collaboration are used to develop new energy services which can unlock the potential of a Smart Grid. The results of the trials are helping to resolve important issues relating to intelligent networks, such as the needs of consumers, new business cases and new laws and regulations. These experiences must give a solid base for strategic decisions on the large scale application of a Smart Grid.

The powermatching concept is considered essential for the scalability of demand response applications that require large amounts of flexible distributed energy resources. Different applications of the PowerMatcher are found in the trials in the Netherlands. Also the large Ecogrid EU demonstration project on the Danish island Bornholm (included in this case book) uses the PowerMatcherconcept.

DNV KEMA, the consortium leader, gives the project also a ‘springboard role’: companies get the opportunity to test their products and services. Next to that is TNO, together with industry partners, developing the PowerMatcher technology into a flexible power platform available for open use in Smart Grid projects worldwide.


Objectives & Benefits

The objectives of the two phases of PowerMatchingCity are different. The objectives of the first phase of this demonstration project were:

  • Demonstrate the feasibility of a Smart Grid / Smart Energy System under real living conditions.
  • Demonstrate an integral optimization method based on local markets and distributed intelligence for both capacity and commodity.
  • Develop an application independent solution.
  • Integrate local generation and demand response.
  • Integrate gas and electricity infrastructure in the most optimal way.

The main objective of the second phase (2011-2014) is to place the technically feasible solution of the first phase in the real energy world:

  • Demonstrate new end user propositions based on real time pricing and energy management insights.
  • Implement the solution into the wholesale processes (allocation, reconciliation) and billing
  • Extend the role of the grid operator: validate the peak load reduction potential by extending the field trial with households behind a single transformer.
  • Extend to a smart electric vehicle charging service.
  • Validate the cost/benefit model with data from the field trial.


Planning for Success

The smart energy system has numerous stakeholders. The concept of PowerMatchingCity is to find an optimization for the various goals of the stakeholders:

Consumers nowadays invest in their own power production, e.g. in PV solar installations. These so-called ‘prosumers’ are looking for the optimal economic benefits of their investments. From a household perspective the network can be regarded as a very large battery. The economic benefits for a prosumer can be maximized by continuously seeking the highest profits for energy export towards the grid and minimizing the costs for import from the grid. This provides the flexible reactive power for a Smart Grid.

Grid operators or Distributions System Operators (DSOs) are confronted with changing energy demands and load profiles. The electrification of the energy system will lead to increased network loads. Extensions of transport capacity of existing networks in especially cities are very expensive and labor intensive operations with a high impact on the built environment. Therefore the development of advanced distribution automation is highly relevant to manage future load profiles and manage congestion and peak loads in local grids and on distribution stations. Within PowerMatchingCity the DSO can influence the load profile on the transformer by giving local price incentives. In this way it can actively limit the import or export of energy.

The interconnection of microCHPs into a Virtual Power Plant (VPP) is now a commonly known concept that can be used for the reduction of power imbalances and for the optimization of trading portfolios. Within PowerMatchingCity the whole system of households and connected devices is treated as a VPP and is directly controlled from the trading room. By continuously altering the balance between energy production and demand the resulting power production or demand of the cluster can smooth peak power demands and prevent the dispatch of costly spinning reserves.

Figure : The PowerMatchingCity projects draw on international experience

For the demonstration of system changing concepts like the PowerMatcher it is essential that they can rely on solid theoretical foundation, connections with other demonstration projects, and scientific analysis of the results. In his thesis ‘The PowerMatcher: Smart coordination for the Smart Electricity Grid’ Koen Kok provides a theoretical basis for the design of the PowerMatcher and an extensive validation through simulation studies and field experiments, including “PowerMatching City”[1]. Some key elements and conclusions in his thesis are[2]:

The design of the PowerMatcher is based on multi-agent systems which makes the system highly scalable and able to ensure user privacy. The theoretical work brings together elements from electrical engineering, computer science, economics and control. Further, it includes a mathematical proof of the optimal performance of the PowerMatcher. (…)

Thus the operation of the electricity system changes from central control of a relatively small number of large power plants to coordination of large amounts of (sustainable) generators and flexible users. An important requirement for this coordination system is scalability. Maintaining the system’s demand and supply balance will involve a huge number of small and medium-sized smart energy demand devices.

Controlling this from a central point will soon reach communication and computational limits. This scalability problem is even greater in the field as the distributed generators will also play a role in the coordination task. Computer science, and in particular the area of multi-agent systems, can offer a solution.

[1]Published by TNO, The Netherlands, SIKSDisseration Series No. 2013-17. Dissertation defended on July 4th 2013, at the VrijeUniversiteit, Amsterdam.

[2]See pages 27r3-276 (Summay).

A multi-agent system is a distributed software system in which so-called intelligent agents are responsible for local sub-tasks, and communicate with each other in order to achieve the higher system goals. A well-designed multi-agent system is an open, flexible and easily expandable ICT system that can properly operate in a highly complex and changing environment. As the local software agents take care of local business, it separates local (and potentially privacy-sensitive) information from the outside world by not collecting it all at a central point.

The PowerMatcher is designed and built based on this multi-agent technology. The result is a mechanism which allows for the coordination of a large number of smaller consuming and producing devices without the autonomy and privacy of the owners of these devices becoming compromised.


Current status & Results

The field trial started in 2007 (the operational part started in 2009) and ended its first phase in 2011.

Phase 1 cost approximately 5 M€ and was partly financed by the EU (PF6).

The trial with 25 homes showed that it is possible to create a Smart Grid or energy network with the associated market model using existing technologies. The system enables consumers to exchange electricity freely and the level of comfort is maintained.

From 2011 to 2014 the second phase is executed. Phase 2 started at the end of 2011 and started the operational phase mid 2013. The cost of phase 2 is also approximately 5 € and is financed with 2 M€ by the Dutch Government. Phase 2 consists of the same 22 households as in phase 1, with additional 18 households, 10 EV’s and 2 smart distribution transformers. The 18 new households are situated in the same street, behind one distribution transformer. The new households are also members of the local energy cooperation that is expanding rapidly with for instance plans for collective solar panels on a local school.

By using the PowerMatcher, more renewable energy may be integrated in the electricity system. A study of the energy consumption of 3000 households in combination with a large (off-shore) wind turbine park clearly shows this. When using the PowerMatcher, it was shown that approximately 65 to 90% of the wind power, which would normally not be used without coordination, could be locally utilised. As a result of this, the usage of power from fossil fuels is reduced by 14 to 21%.

A reaction from energy demand and distributed generators to fluctuations in the supply of renewable energy also improves the value of green power. The low day-ahead predictability of wind generation, for example, results in additional costs assigned through the electricity wholesale markets, the so-called imbalance costs. In two of the field tests performed with PowerMatcher, a wind farm was coupled to a flexible cluster in order to compensate for deviations from the wind powerprediction. This reduced the imbalance caused by the wind farm 40 to 60%. This makes an interesting business case for energy suppliers.

Further, it has been shown in the field that the PowerMatcher is able to avoid overloading of electricity networks. By cleverly managing heating systems (micro-CHP and heat pumps) and/or charging electric cars, the daily peak loading could typically be reduced by 30 to 35%. In existing networks, this saves the network operator an expensive network reinforcement, while new networks can be less heavy designed. In one of the cases studied, the network capacity could be designed three times lower through application of the PowerMatcher.

Results of field tests in PowerMatching City

The results of the field test show that the market control mechanism works perfectly. The smart agents ensure that end users buy their electricity at low prices and sell at high prices (see example below). The tests also show that the cluster can be operated as a virtual power plant and grid operators can send incentives to reduce the peak load in certain areas of the grid.


Figure: Example of the reduction of solar power imbalance in PowerMatching City: (top-left) a) Sun Imbalance in development: Prediction , 
Realization and Imbalance, (top-right) b) Real time price development, (bottom-left) c) Collective response micro cogeneration units systems, 
(bottom-right) d) Collective response smart hybrid heat pump systems

It is technically feasible to allow demand response to track supply rather than the other way around, as is currently the case. Measurements from the micro-CHP, the hybrid pumps and the charging of electric vehicles all indicate that the system responds quickly to fluctuating demand and maintains comfort levels for the end user over the long term. This is favorable for the smooth integration of renewable wind and solar energy.


Lessons Learned & Best Practices

A working Smart Grid system as starting point

The goal of the PowerMatching City project is to demonstrate that a smart system for the future supply of energy can be built using readily available technologies. Phase 1 has succeeded in doing this. The connection of different energy flows was successful. The intelligent combination of electrical vehicles, micro-chp and heat pumps caused lower peaks in the grid and to work as a virtual power plant. The system works well, although not without applying the necessary creativity. The main lesson from the first phase is that the chosen solution is a technically feasible solution. However the results from phase 1 make it difficult to state the reduction of the energy use.

A practical trial like this turns out to be extremely well-suited for acquiring insight into what can be achieved with a smart energy system, the changes required for this purpose, as well as the hurdles still need to be overcome. One of the most important lessons to be drawn from Phase 1 is that it is only through the efforts of all parties along the entire energy chain that it becomes possible to fully exploit the opportunities inherent in a Smart Grid.

The challenge

In the future, to ensure the security of electricity supply a new coordination mechanism is required. The reliability of our electricity supply will need special attention due to three developments:

  • The rapid increase in renewable energy creates a challenge for maintaining the crucial balance of supply and demand in the network.
  • The growing use of electricity, which increasingly drives aging networks towards overload.
  • Part of the electricity generated is becoming distributed: large numbers of relatively small generators –  solar panels, small wind turbines and micro-combined heat and power – deliver their energy close to the place of consumption. These generators operate outside the reach of the central coordination within the electricity system.

Smart appliances

From actual practice it becomes evident that there is a need to design appliances, including house-hold appliances, differently. Appliances should be allowed to decide for themselves whether to switch on or off, depending on the current electricity rate, for example, when the rate falls because the supply from renewable sources is high. This implies that the devices have knowledge on the electricity rates, communicated via the internet. In PowerMatching City we have adapted the heat pump, microCHP, EV and washing machines in such a way that they are able to communicate with the Smart Grid. Making these interfaces is not always trivial.

The trick is to create flexibility without adversely affecting the end user’s comfort or the system’s energy efficiency. A heat pump, for example, has been designed to supply heat when the consumer has a need, not when the electricity rate is favorable. To make it flexible is possible by temporarily storing the energy in a buffer tank in the form of heat. The battery in electric cars offers similar potential. By charging the battery at a time when electricity is cheapest, it is possible to drive the car at the lowest possible cost. The project demonstrates that each of the innovative technologies developed for this project provides a significant amount of flexibility and can be operated flawlessly.


Households perceive a high level of comfort and don’t experience any inconvenience from participating in this Smart Grid project since energy trading is fully automated. The acceptance level is high, and a clear change in ‘energy behavior’ can be observed. The effect on their direct energy consumption is limited, but the end-users show an increased willingness to invest in more energy efficient appliances once evidence is provided that their investments result in the expected savings. Moreover the participants gradually transform into prosumers and more and more they want to (collectively) generate their own power.

New elements in phase 2

Three important new elements in phase 2 are new energy services, addressing the problem of network capacity management and the ‘springboard role’.

A lot of attention has been given to the end-user requirements and desires with regard to new services and the systems to deliver these services. These services and a home energy management system with a tablet have been developed with intense customer consultation. One question is how intelligent networks can be incorporated into the energy company’s processes, from reading the meter through to billing. For example, there are tariffs which can change every 5 minutes instead of a fixed tariff. That requires new sorts of bills.

Households get their bills every month, based on their actual electricity consumption. Two types of contracts have been developed, after extensive consultation with the participants:

  1. Cheapest energy bill: The PowerMatcher will control the smart devices in such a way – within the comfort levels as demanded by the households : that way the energy bill will be as low as possible.
  2. Sustainable local first: The Powermatcher will control the smart devices in such a way – within the comfort levels as demand by the households – that locally produced energy will be used to a maximum.

Secondly PowerMatching City II is also addressing the problem of capacity management: how can you ‘feed in’ large quantities of renewable energy into the network?

And third, the project has a ‘springboard role’: companies have the opportunity to test their products and services. Hence NXP will be building computer chips into electric scooters in order to be able to charge them smartly. iNRG will be experimenting with the communications between a HRe boiler and water heater tank, which makes it possible to break the link between the demand for heat and electricity production. These are new steps on an already working Smart Grid and an infrastructure.


Key Regulations, Legislation & Guidelines

In the existing Dutch framework of energy laws and corresponding codes SMEs and domestic end-users are characterized by demand profiles, which are used for settlement, allocation and reconciliation. The introduction of active demand and supply within the SMEs and domestic end-users will result in shifted load profiles as a result of optimizing supply and demand of energy. These shifted load profiles will deviate from the statistically determined load profiles that are currently used to determine the average load profile of these groups of end-users. The flexibility that is created this way can have a value in the energy chain: not only by preventing congestions in the network and giving trade options on the commodity market, but also by giving balancing options. But this requires that the ‘standard load profiles’ be set aside. Therefore the value cannot be assigned to the corresponding parties who created it in the energy system.

In PowerMatching City we explore how these wholesale processes can be adjusted such that the value created by the flexibility introduced by active demand and supply can be valorized allowing that parties responsible for balancing the system can share this value with the associated end-users.

Next to adjustment of the profile methodology the current tariff structures most likely need to be adjusted in the near future to fully unleash the potential of a Smart Grid and accelerate the introduction of local renewable energy production by residential end-users and SMEs.

  1. Within PowerMatching City value for flexibility is created with real-time local tariffs. To valorize the flexibility the standard profiles methodology needs to be changed and dynamic tariffs both for the commodity as for the capacity are needed. Such a market model would require adjustment of the existing tariff regulations but allow for transparent cost and benefit allocation, an optimal dispatch of all assets in the system as well as freedom of dispatch, transaction and connection.
  2. Currently the feed-in tariffs in The Netherlands are capped at 5000 kWh/y. Above this threshold the benefits for feeding power in the grid reduce dramatically. When the drop in prices from renewable sources like wind and solar power continues and production volumes increase they will significantly shift the energy prices as a result, a fixed feed-in tariff, soon will no longer be socially acceptable or technically feasible.


back to top