By Jacob Klimstra, Wärtsilä, the Netherlands
The use of electricity and human activities are, of course, closely connected. During the night, the demand for electricity falls to a minimum, but from 7 am on weekdays many activities start up again, resulting in a sharp rise in power demand, which can be up to 180 per cent of the minimum demand. In hotter countries, air conditioning causes a peak in demand at in the afternoon. In colder countries, demand peaks in wintertime can be 50 per cent higher than in summer as electricity is used for domestic heating.
Network operators need to be able to control the output of the generators in a supply system to cope with these short-term fluctuations in demand. These generators can then be used for frequency regulation, load following and as spinning/non-spinning reserve. Network operators predict the demand pattern for the next day and contract sufficient capacity to meet the sharp rise in power in the morning and the peak in the afternoon.
Steam-based power plants typically need a couple of hours for preheating before they can deliver electricity. Their ramping-up capacity is about two to three per cent of their nominal power per minute. Power stations can also fail to start up or trip at full load. That is why a contingency reserve is required, which can be spinning reserve (running and on-line) as well as stand-by reserve (non-spinning, but ready to come on-line).
The non-spinning reserve power has to come on-line as soon as the bulk of the spinning reserve has been used. It will be clear that a generator owner has to be compensated financially for such so-called system, or ancillary, services. Reciprocating engines and aero-derivative gas turbines have a faster ramping-up rate than steam-based power systems and so are often used for this type of services.
Electricity production characteristics of renewables
Electricity production from wind energy depends on the local weather conditions and is largely not controllable (non-dispatchable). If the wind increases in strength, there is a high risk that the wind turbines will suddenly have to cease operation while running at rated capacity, to avoid damage caused by overspeed. To cope with these situations, the network needs spinning capacity that can instantaneously take over the load from the wind turbines.
Also, if the wind speed falls drastically over a wide area, much stand-by reserve has to be put on-line. In both cases, the network operator has to pay for these services, which can reduce to virtually nothing the intrinsic economic value of the electricity, therefore the real economic value of the wind energy produced depends on the extent to which the network operator can use it to match demand.
Biomass-based power plants differ in this respect, however, since their fuel can be stored close to the power plant, making the output more dispatchable. Liquid biofuels are ideal for diesel engines that can act as spinning and back-up reserve for electricity from wind.
Wind speed has a stochastic character. The average wind speed at offshore locations is normally higher than that at land-based sites, giving a higher capacity factor.
The capacity factor of a wind farm is the total amount of electricity produced in a year divided by the electricity that would be produced if the generator was running 100 per cent of the time at full load. Figure 1 illustrates typical production duration curves for a wind turbine at an average site and at an excellent site. In addition, a wind turbine will also have downtime for maintenance.
Figure 1: Typical output during a year for a wind turbine at an average site (red) and at an excellent site (blue).
A new offshore wind park of 110 MW near the coast of the Netherlands (Shell/NUON) is expected to have a capacity factor of 30 per cent. The specific capital investment in that park is €2100 ($2840) per kW. For a commercial fixed interest rate of 10 per cent, the specific capital costs will be €0.08 per kWh. Insurance costs as well as operation and maintenance costs will easily result in production costs of more than €0.11 per kWh. That is rather high for non-dispatchable electricity. Land-based wind installations are considerably cheaper than offshore installations (€1200-1500 per kW), but their capacity factor is generally just slightly above 20 per cent.
The Danish dilemma
The country with the highest relative share of wind energy based electricity is Denmark. In 2004, 16.3 per cent of its total 40.5 TWh of electricity was produced by wind power. However, the capacity factor for the total installed wind power was just 24 per cent.
Assuming that on average 5 per cent of the wind turbines were not available for production because of maintenance and repairs, the available wind capacity was 0.95 x 3.1 GW = 2.9 GW. Total annual net electricity generation from all sources in Denmark is 38.4 TWh, which means a time-averaged production of 4.4 GW.
Wind power in Denmark will exceed the nightly minimum weekday demand if the 2.9 GW active wind capacity is on-line with a favourable wind. With a policy of unlimited grid feed-in by wind power, the only solution in this situation is to export. That explains why Denmark has relatively high electricity exports compared with typical EU-27 countries.
The problem of having too much electricity production on-line is exacerbated in winter in Denmark since the demand for heating is often met by combined heat and power units. That means that, especially during cold windy nights, even more excess electricity is produced. If there are high winds, much fuel-based or hydro-based spinning reserve and back-up power has to be available in case the critical wind speed is exceeded. This is quite costly and results in spoiled fuel consumption.
On the other hand, wind speed levels are low for a significant part of the year, so power plants based on fuels have to take over. As a result, the utilization factor for conventional thermal plants in Denmark is only 37 per cent, while the general optimum for the sector lies between 50 and 55 per cent. Denmark also actively uses electricity imports to cover peaks in demand (see Figure 2). Denmark itself has no pumped hydro storage but can use Norway’s storage capacity to some extent.
Figure 2 : An illustration of Denmark’s relatively high electricity imports and exports due to substantial utilization of wind power.
Without wind power that can freely feed into the grid, nuclear and coal-based power plants normally cover the baseload power generation so that capacity factors of up to 85 per cent can be reached. However, with a high amount of wind power such as in Denmark, nuclear power is especially not attractive since it would have to be shut down during nights with much wind. The nuclear process is not suitable for this sort of operation, and it would also substantially increase the specific capital costs of electricity.
In summary, the stochastic character of wind energy makes it necessary to install additional dispatchable and flexible back-up power plants that will run with a low utilization factor. Electricity in Denmark for domestic users costs €0.24 ($0.33) per kWh in 2004, twice as high as the EU-27 average. To be able to produce 16.3 per cent of all the electricity used by wind, the wind capacity had to be 23 per cent of the total capacity, while the total installed generation capacity had a utilization factor of only 34.6 per cent.
A hybrid solution to effectively integrate renewables
The basic character of wind power means that the installed wind capacity cannot be considered as controllable, and sufficient alternative capacity is needed to at least cover the peaks in demand. Simple load shedding during conditions when the wind farms cannot produce is generally not acceptable.
Since wind capacity reduces the utilization factor of non-wind capacity, it is necessary to have a back-up capacity with a low specific investment. In addition, the non-wind generators need to have high ramping up and ramping down rates. That is required at times when wind power covers the baseload (a maximum of 20 to 30 per cent of the time), since then the other generators have to take care of the rapid rise in demand associated with the intermediate load on weekday mornings. A rapid response is also required in case the wind power suddenly stops because of excessively high winds.
Wärtsilä 20V32 engine with 9200 kW output, an engine that can also run on liquid biofuel.
Fossil fuels are certain to become scarcer during the planned life of at least 30 years for future generating equipment, so high fuel efficiency is very important. Power plants consisting of generators driven by reciprocating engines can have a simple-cycle electrical efficiency of up to 45 per cent, can ramp up to full power in two minutes (ramp-rate 50 per cent per minute) and start from standstill to full load within ten minutes.
Such units also have the advantage of a quite flat efficiency curve in the upper load range, which is attractive for offering spinning reserve. Moreover, the specific investments for gas-fuelled installations are only about €500 per kW. Installations running on bioliquids can cost up to €700 per kWh. An attractive option is the use of dual-fuel engines. Such engines can run on renewable liquid fuels if these are available and switch over to liquid or gaseous fossil fuel when renewable fuels are not available.
Liquid biofuels are based on rapeseed, jatropha, palm oil or animal fats. Such fuels can easily be used with a high efficiency in reciprocating engines with a power capacity of up to 18 MW. The specific capital investment in such decentralised installations is about the same as for large-scale generators.
The small-scale character of these installations means that electricity can be produced close to the users so that the heat released can also be used (cogeneration), resulting in total fuel efficiency exceeding 85 per cent. The starting time of such installations is less than 10 minutes, so their output can serve as back-up power in the electricity supply system.
Steaming to victory
The electrical efficiency of such installations can be further improved to roughly 50 per cent by adding a steam cycle or organic Rankine cycle. Until now, the additional investment of about €1200 per kW for such a topping cycle has been considered uneconomic. However, compared with wind power, it is quite attractive since here again the ‘fuel’ is free except in the case of high-temperature cogeneration. The capacity is also more controllable than that of wind power.
As mentioned before, using substantial wind capacity will reduce the utilisation factor of other installed generating capacity. That, in turn, will result in higher specific capital costs for the non-wind capacity. In an honest comparison, these extra costs should be attributed to that of electricity from wind. For gas engine driven installations, the specific capital costs of electricity production will increase from €0.01 per kWh to €0.015 per kWh if the utilisation factor decreases from 55 per cent to 35 per cent. The difference would be at least three to six times higher for coal-fuelled and nuclear power plants (see Table 1).
Power stations based on reciprocating engines have multiple units that run in parallel. That guarantees high reliability and availability. By using engines with a unit power capacity ranging between 5 MW and 18 MW, the size of such power stations can easily match the output of wind farms.
The power stations can be built at suitable locations, preferably close to heat users for cogeneration. Generating units at different locations can even be combined into virtual power plants. The units based on reciprocating engines should be equipped with heat recovery as much as possible. Heat storage systems may be needed to make sure that sufficient heating capacity is available in case the cogeneration units cannot run during cold nights because much wind capacity is on-line.
In conclusion, liquid biofuel-fired and gas-fuelled power plants based on efficient and flexible engine-driven generators, with rapid starting and ramping up capabilities, as well as a flat efficiency curve in the upper load range create cost-effective and energy-efficient reserve power for wind farms.