*If higher feed-in tariffs result in a surge in microgeneration, Dan Lewis warns that local electricity networks will struggle to cope.*

Some would have us believe that we are on the cusp of a decentralised micro­power energy revolution, but could the distribution network cope with large amounts of intermittent micro-power?

Today, the impact of distributed micro­generation on Britain's overall electricity supply is tiny. Out of a combined capacity of around 80GW, small wind contributes just over 20MW, small hydro (less than 5MW) around 70MW, solar photovoltaic 22MW and combined heat and power 214MWe, for plants smaller than 1MWe.

*Feed-in tariffs*

There is, however, no question that more microgeneration is coming, because from 1 April 2010, feed-in tariffs rise from 10p or less per kilowatt hour to 36.5p/kWh for small solar photovoltaic systems up to 4kW, 28p/kWh for systems up to 10kW and up to 23p/kWh for small wind turbines between 1.5kW and 15kW. This raises yield on investment by between 5 and 8 per cent, according to the government, reducing payback time and making it more of an investment rather than a lifestyle choice.

In preparation for the changes, the Energy Networks Association has set up an Energy Networks Futures Group with nine working groups to look at the technical challenges that networks face - and there are some real technical challenges. No developed country has run a system with very high levels of distributed generation. Some are predicting unpleasant surprises in store.

*Embedded generation*

Particularly difficult are the network implications of increasing embedded generation. Probably the leading expert on this is Michael Laughton, emeritus professor of electrical engineering at the University of London. In an October 2006 paper supported by British Energy, he cited a wide range of issues which show we are a long way from the plug and play smart grid that many aspire to.

For a start, as the number of embedded generators increases, voltage loads will start to be in excess of transformer loadings. Fault levels will rise, necessitating an upgrade programme - for example, to increase circuit breaker ratings

and reconfigure the network, or for sequential switching.

*Cost variations*

Solutions and costs will be necessarily highly localised. One part of the distribution network may be endowed with excess headroom capacity and have modern transformers designed to cope with an increase in demand and generation. In other parts, the local network may be low voltage and close to its limits. In each scenario, the cost of connecting the same microgeneration technology would vary widely.

Today, as it stands, if you connect less than 16 amps per phase, then you do not need to contact the distribution network operator before you connect up your micro-power station, although you must do so within 28 days. If it is bigger than that, you must contact the network operator before you connect. It will then analyse the local connection and come back with a charge for it. As embedded generation rises, this sort of tricky and unquantifiable decision could become even harder.

Laughton also cites the need for careful engineering and control to minimise the effects of voltage dips, voltage spikes, interruptions in supply and what is known as harmonics - supply ripples that are multiples of the fundamental 50Hz supply frequency.

*Active networks*

He concludes that a solution would be to move the distribution network to an active network environment - one that can dispatch power, be capable of reactive power management and one that can control voltage. Since 2006, though, two known unknowns have started to appear: smart meters and electric cars.

Smart meters have had a very good press and supporters would argue that they are vital to the creation of an intelligent distribution network that can cope with large quantities of intermittent power. At an estimated fully installed cost of £150-£200, they will not come cheap, though, and will undoubtedly benefit distribution network operators more than consumers.

*Electric cars*

In the longer term, say from the late 2020s, plug-in electric cars or range-extended hybrids could be connected to the grid when not in use and used alternatively as backup for load-shaving and for energy storage in times of surplus cheap electricity. With 33 million vehicles on the road today and a typical mid-sized engine size as much as 65kW, this really would be a revolution. Just one million plugged in vehicles could equal 65GW. And yet, exciting as this might be, there is not a single electric vehicle operating in this manner and feeding power in to the distribution network, because the infrastructure does not exist.

In these straightened times, we must be prepared to be sceptical and ask hard questions about the costs and benefits of micro-generating distributed energy.

Dan Lewis is chief executive of the Economic Policy Centre.

*Definitions*

Microgeneration can be broadly defined as the small-scale production of heat or electricity from a low carbon source. Typically, this would include solar photovoltaic panels, small wind turbines and micro-combined heat and power. Overlapping strongly with microgeneration - because no-one really agrees what the capacity cut-off is - is decentralised energy. This is energy that is generated at, or near to, its point of use and does not rely on the high-voltage transmission system, but on the distribution network. This may include energy generated for a single home or for a local community.