Our energy networks are facing a challenging time. Their infrastructure is generally ageing. Yet at the same time we are asking them to work harder than ever before. In the past, networks were required only to handle a simple, passive flow of power from high-voltage generation and transmission to low-voltage consumption. Now they must make the transition to handling complex, highly variable and multi-directional power flows to accommodate:
- increased levels of distributed generation
- the potential transition of energy sources currently on the heat grid on to the electrical grid (for example, ground- and air-source heat pumps)
- the growing demand for electric vehicles.
One of the greatest challenges for the network is the inherently variable nature of distributed renewable energy sources such as wind and solar power. Also, whereas power stations burning fossil fuels are normally located conveniently close to the load centre, the location of wind farms is dictated by meteorological considerations – they have to be constructed where the wind blows. So, in many cases, the point of generation is separated from the point of consumption.
An additional concern is that renewable generation often has essentially low or zero inertia. In our quest to de-carbonize power generation, it is not enough to just add renewable sources to the mix. We need to decommission thermal power stations, or at least reduce their run-time. But that significantly reduces the inertia present in the form of the rotating machinery, which is in itself really just a form of energy storage. Without this inertia, the network becomes inherently more unstable and sensitive to power fluctuations and mismatches between demand and supply.
The three factors outlined above – more variable generation profile, more variable and complex demand profile, and lower system inertia are coming together to create a ‘perfect storm’ of challenges facing our future networks. However, the deployment of energy storage can offer a shelter against these concerns that is extremely attractive for technical reasons, commercial reasons and electricity market interaction reasons.
Frequency response and control
In the UK, frequency response and control is the responsibility of National Grid. Currently it does this by buying services from generation companies. But if energy storage devices are deployed they could provide this extremely valuable service.
Supporting energy markets
Another aspect where energy storage offers exciting possibilities is in the energy market of the future. Simulations projecting weather conditions and market penetration of renewable generation to 2030 indicate that a significant differential in the price of electricity will arise depending on the availability of generation sources. The price in £/MWh could range from around zero, when the wind is blowing strongly, to £2,000 when it doesn’t blow at all!
The emergence of this margin, which isn’t really a factor in the current market, offers an opportunity for operators with energy storage systems to charge them when the price is low and to sell the power on when the price is high – presenting a powerful business case. This is often referred to as arbitrage.
Energy storage can also play an important role in the electricity market by helping to close the discrepancy between actual and contracted volumes of energy purchased by supply companies. This can remove a level of uncertainty as well as saving the costs of having to go to the balancing market.
Distribution Network Operators (DNOs) could usefully employ energy storage to defer, or perhaps even eliminate, the need to make major investments in their network infrastructure to handle demand peaks. So when a demand peak occurs – with sharp peaks likely to become more frequent with the advent of electric vehicles – energy storage will provide the necessary support.
Voltage control, to maintain network voltage within statutory limits becomes more difficult on a network that has sharp peaks as well as a high level of distributed generation. Energy storage could be an important aid to network voltage control, not just with real power at lower voltage, but also when used in conjunction with sophisticated power electronics – such as in ABB’s DynaPeaQ® concept – it can provide full four-quadrant operation, including reactive power.
In networks that are being run very hard, requiring a lot of reactive power, losses could be reduced. This would be achieved by using energy storage to meet reactive power requirements, as well as chopping off peak power consumption, hence reducing I2R losses.
If energy storage is deployed at the optimum, strategic point in the local network, it could play a very significant role in post-fault restoration. Where a fault condition has occurred, it is generally possibly to get between 60 to 90 percent of customers back online fairly quickly by rerouting. But voltage and thermal limits restrict how many customers can be restored. An energy storage device could help to prop up the network both thermally and in terms of voltage, keeping it buoyant with a higher number of customers restored post-fault. This would contribute to a reduction in customer minutes lost (CMLs), which is a key performance indicator for utilities.
Switching between feeders
If an energy storage device is positioned at a normally open point (NOP), it is possible to switch it between feeders. The benefit of this approach is that not only can the system switch between feeders as required, it can use them as part of a complementary charge and discharge cycle. So for example, if feeder one is lightly loaded and feeder two is heavily loaded, the system can ‘hop’ between the two – charging from feeder one and discharging to support feeder two.
It is important to note that it is possible to model very effectively the potential impact of energy storage on a network. Durham University has developed a dedicated Energy Storage Evaluation Environment for this purpose.
When designing energy storage installations it is vital to consider where they will be positioned in the network as well as their size in terms of rating of the power electronics and the capacity in Ah. Furthermore, it is essential to have a fully developed implementation strategy: if energy storage is used recklessly, it will not provide the anticipated benefits and will age prematurely.
It is not always necessary, or desirable, to provide a very large energy storage system with real power output measured in megawatts. There are cases where it is more appropriate to provide a 400 kW intervention for a sustained period, rather than 1 MW for a shorter period.
As distributed generation grows in networks, there is an increasing possibility for islanding of systems. Consumers might well ask why they have suffered a loss of supply resulting from a fault several miles away when they have local generation? An islanded network can run independently until the fault is cleared and then resynchronize with the grid. But when the island is isolated it loses the frequency and voltage support from the grid. This is where energy storage comes in to provide the controllable real and reactive power to maintain stability.
Distributed generation – wind farms
Energy storage can also play a role for operators of distributed generation. Consider the case of a wind farm operator that would like to build a 15 MW wind farm, but finds that conditions imposed by the network operator concerned about voltage or overpower problems would restrict this to 10 MW. By building a 15 MW wind farm in conjunction with energy storage, the operator has the contingency to divert the excess power into storage when a network constraint arises, and then feed it into the network when the constraint is cleared.
A further possibility is to run the energy storage continuously alongside the wind turbines, acting as a shock absorber to provide smooth output power. This smoothed and more predictable supply is ultimately more valuable to the customer and commands a better price.
Some older designs of wind turbines such as the directly connected induction generator (DIG) absorb reactive power that varies according to how much real power they produce. An energy storage device can dynamically compensate for this active power demand.
Fault ride-through is an important design consideration for wind farms. In cases where a short-circuit occurs on the grid and the voltage drops, the Grid Code dictates that turbines must stay connected to support the grid and must also pump in Volt-Amperes-Reactive (VARs) to support the voltage.
The problem for wind turbines is that if the voltage drops the mechanical load on their blades has nowhere to go, so there is a risk of them speeding up and losing stability. But with energy storage, they can dump the excess energy into it, using it to maintain stability for the very short ride-through period of 150 to 200 msec until the network protection systems operate and the circuit breaker opens.
The investigations carried out at Durham University to date have concentrated mainly on the potential impact of energy storage on networks. The conclusions show clearly that energy storage could make a very positive contribution in many areas of the network. The next steps are to put theory into practice and some test installations are already under way.