Commercial Measures to Reduce the Cost of Wind Integration in the Island of Ireland

Guest post by Riccardo Carollo of Incoteco ApS

This report provides a comprehensive description of the Irish electricity generating system and how it is evolving to cope with ever higher levels of intermittent wind power. Of particular interest, the report contains information on actual generation for specific power stations and shows how their use has declined between 2010 and 2015. The report also describes commercial / technical solutions to efficiently deal with the load balancing issue. Neither I nor Energy Matters have any commercial relationship with the companies involved: Incoteco ApS, Rolls Royce plc or Ormat inc.

Executive Summary

  • The island of Ireland experienced a rapid growth in the wind power installed capacity over the last years, with more than 3 GW all-island capacity at the end of 2015. This trend is expected to increase. Both the Republic of Ireland and Northern Ireland have agreed a target of 40% renewable share in the fuel mix by 2020 – the vast majority of it coming from onshore wind power.
  • Ireland is the electric island with the highest share of wind power generation worldwide. Because of the scarcity of interconnections with other power systems and the relatively small scale of its pumped hydro storage system, the existing thermal fleet has to absorb wind power’s intermittent output. The change in the operation of the Irish CCGT fleet, in particular, has been dramatic over the last years. Since 2008, the fleet’s capacity factor has halved, falling to only 40% in 2015. Most CCGTs are now online only during peak load demand and/or when wind power output is very low or null.
  • The CCGT fleet was designed for base-load operation, achieving high efficiencies (over 50%) at constant rated power output. The flexibility required from these units to integrate an increasing share of wind power has resulted in dramatically increased O&M costs, specific fuel consumption and specific CO2 emissions.
  • This study examines in some detail, how the system operator is forced to cycle even “base-load” stations and call on other CCGTs that were designed as base-load units to operate intermittently as “peakers” and in-fill power suppliers from the SEMO database.
  • Incoteco (Denmark) ApS has commercial involvement with two possible commercial solutions for the Irish case.
  • Eos Energy Storage ( has developed modular MW scale battery storage modules, which can be installed rapidly in virtually any grid or distributed connected location, whether at substations or directly at the wind farm site. At an ex-factory price of 160 US$/kWh (640 US$/kW), this technology is lower in cost than all its known competitors. The installation of fast-acting storage systems would absorb part of wind power’s fluctuating output, allowing the CCGT fleet to operate with higher efficiency and reduced costs. If installed at or close to the wind farm site, battery storage can also help reduce growing wind power curtailment. Curtailment has been increasing in direct proportion to the wind power installed capacity.
  • The second solution Incoteco presents for this dilemma is the Trent-Ormat combined cycle gas turbine. In modules of 134 MW, this technology combines relatively high efficiency (46% HHV) with high flexibility.
  • The gas turbines synchronize in 5 minutes and run to full load in another 5, while the rated power (gas + organic Rankine cycle turbine) synchronizes in only 5 minutes and reaches full output within 15 minutes after “pushing the button”. In the event of the decommissioning of one or more of the coal-fired thermal units, the replacement of the installed capacity with Trent-Ormat turbines would allow higher system flexibility and relieve the existing, high efficient CCGT fleet from absorbing much of the wind power’s intermittent output.

1) State of the Irish Power System in 2016

Ireland is an electricity island, having limited HVDC interconnections with Great Britain only, the 500 MW East-West Interconnector, commissioned in 2012, and the 500 MW Moyle Interconnector. New installations, such as the proposed Celtic Interconnector linking Ireland and France, are not expected be built before 2020 when wind penetration is supposed to be 40%.

Due to the effects of the 2008 economic recession and because of the increased efficiency in the electricity supply, the Irish energy demand has not increased during the last ten years [1].

Figure 1: Primary Fuel Mix for Electricity Generation for the Republic of Ireland [1].
The stagnation of the island’s energy demand can be seen from the plot of Figure 2. Over the last years, the Irish yearly electrical load demand has been around 35 TWh/y.

Figure 2: Annual Electricity Consumption in Ireland. Sources: [20], [21].

Wind power installed capacity has been growing. The island now has one of the highest share of wind power in its fuel mix worldwide, and the highest for a large electricity island. The trend can be seen from the plot of Figure 3. In 2015, Ireland generated over 23% of its electricity demand from wind power, which has over
3 GW of installed capacity [2].

Figure 3: Installed Wind Power Capacity in the Island of Ireland [4], [5].

Both the Republic of Ireland and Northern Ireland have agreed the ambitious target for 2020, when 40% of the electrical energy production in the whole island is expected to come from renewable energy sources, mainly onshore wind power. It is not clear whether Ireland will succeed in reaching the 2020 target, since this depends on other factors than the installed capacity such as the capacity factor of the new installation sites, the system load demand and the wind power curtailment due to system limitations.

What does seem certain is that wind power’s installed capacity and the wind share in the fuel mix will continue increasing during the next years. Due to the high intermittency of wind power generation, and the above-mentioned lack of interconnections with neighboring power systems, the grid integration issues studies in this report will increase.

Ireland lacks significant storage systems apart from the 292 MW Turlough Hill Pumped Storage, a unit that is used to provide primary and secondary operating reserve [3]. For these reasons, wind power intermittencies have to be mainly absorbed by the existing thermal units installed in the island.

In January 2016, there were 8 CCGT power plants registered in the SEMO database, 7 in the Republic of Ireland and 1 (Coolkeeragh CCGT) in Northern Ireland. CCGT total installed power amounts to 3.7 GW, with the individual plants’ rated power between 384 and 747 MW. All of the CCGT units are relatively new, having been commissioned between the year 2000 and 2014. These plants were originally designed for base-load operation, achieving high efficiencies in a short range around the design power output.
In addition to the CCGT fleet, there are two coal-fired power plants within the Irish system. These the 3-unit, 915 MW Moneypoint Power Station in SW Ireland and the 520 MW Kilroot Power Station in Northern Ireland. Three peat-fired plants, with overall rated power of 370 MW, are distributed around the Republic and are used to provide base-load generation. Hydro’s contribution to the fuel mix is limited, due to the limited installed capacity of 237 MW (excluding Turlough Hill pumped storage). The 640 MW open cycle gas turbine fleet complete the thermal fleet; these units are activated mainly during peak load demand [6].

2) The Irish Electricity Market

The Single Electricity Market Operator (SEMO) operates the all-island electricity market for Ireland.

All electricity is centrally traded through a pool system, where licensed generators sell their electricity to a licensed supplier who sells it onto the pool, and receives a single market price (SMP). The electricity market prices are set by the SEMO, and they are published on half-hourly basis and calculated after the event.

In addition to the energy payment, the market price per MWh sold, the owner of a Generating Unit may access two additional streams of revenue. These are the capacity payments – compensation for being available to generate upon instruction from the grid operator – and constraint payments – compensation for being constrained from exporting the scheduled amount of energy onto the system because of network or system limitations [19].

Over the last years the capacity and constraint payments for the All-Island system have been constantly over 500 and 100 millions of € respectively. These data can be found listed in Table 1. Figure 4 illustrates the trend in the yearly payments for the All-Island market (values in millions of €).

Table 1: Yearly payments for the Generating Units in the SEMO over the last years. Values in millions of €. Source: SEMO [8].
*Partial value for 2016, updated to April 2016.

Figure 4: Yearly Payments for the All-Island Market. Source: SEMO.

In the next years, the capacity payment mechanism is expected to change. According to the 2016 EirGrid “All-Island Generation Capacity Statement 2016-2025” [6] the introduction of a new Capacity Remuneration Mechanism (CRM), will be a major change for all generators. This system is expected to be operative from 2017. Until these new mechanisms can be understood by potential investors in electricity storage or alternative technologies to deliver more flexible responses to wind’s stochastic intermittency, it will not be possible to make a business case for the introduction of these.

3) CCGT Present Operation

As mentioned in the previous section, all the power plants in the Irish CCGT fleet were originally designed to perform in base-load operation. The increase in the wind power installed capacity, combined with the stagnation of the load demand, has resulted in highly variable operations for all the CCGTs, which are now primarily responsible for absorbing wind power’s constantly fluctuating power output.

As it can be seen from the All-Ireland Fuel Mix during November 2015, plotted in Figure 5, CCGT energy production is highly dependent on the wind power output. During periods of low wind power production, such as the first week of November 2015, the CCGT fleet is used to cover the majority of the load demand. During periods of high wind production, wind supplies up to 46% of the demand [7] and the CCGT fleet is used with a low capacity factor. Coal and peat units are normally used to provide base-load generation.

Figure 5: All-Island fuel mix during November 2015. Calculations by Maria Tsagkaraki.

In this work, the data available from the Single Electricity Market Operator (SEMO) database are used to determine the operation of the generators in the island [8]. With a detailed analysis on the single plants’ production within the CCGT fleet, it can be seen that Dublin Bay, Coolkeeragh and Whitegate CCGTs are providing most CCGT generation – although their power output also fluctuates according to the net load demand. The remaining units in the fleet – Aghada, Great Island, Huntstown, Poolbeg and Tynagh – are instead used to meet the peak and fluctuating load demand with rare episodes of higher output during periods of low wind production. This is visualized in Figure 6.

Figure 6: CCGT Power Output during November 2015. “Other CCGTs” includes Aghada, Great Island, Huntstown, Poolbeg and Tynagh CCGT. Source: SEMO.

The number of start-ups and stops and the load following cycles have increased during recent years. As examples, the power output of Tynagh and Huntstown Power Stations, have been analyzed for the years 2010 and 2015. The analyses are shown in Figure 7 and Figure 8.

Figure 7: Power output of Tynagh CCGT in 2010 and 2015. Source: SEMO.

Figure 8: Power output of Huntstown CCGT in 2010 and 2015. Source: SEMO.

The SEMO database was analyzed for November 2015 when during the first week there were particularly low levels of wind power production in Ireland. The power output of each of the 8 CCGT plants in the Irish fleet is shown in the charts of Figure 9.

Figure 9: Power outputs of each unit in the Irish CCGT fleet during November 2015. Source: SEMO.

This flexible operation greatly affects the efficiency of the CCGT fleet, as demonstrated in the report “All Island Specific CO2 Emissions CCGT Fleet and Wind Penetration 2014-2015.” by Maria Tsagkaraki [9]. The CCGT fleet’s fuel efficiency during recent years has fallen far below the nameplate value of the individual generators, amounting to 32% HHV in 2014 and 34% HHV in 2015. Increased specific-fuel consumption of the fleet is caused mainly by the growing number of start-ups, and because of the lower efficiency when the plant is operated below the design rated power as clearly shown in Figure 7, which is typical for all the 23 months analyzed in Maria Tsagkaraki’s report [9].

Therefore the flexibility required from the CCGT fleet comes at a higher cost than widely believed, as explained in the next section of this report.

4) An Evaluation of the Cost of CCGT Cycling

The culmination of adding more variability and unpredictability to a power system, as the share of wind power increases, is that, lacking efficient storage, the thermal units will be required to start-up, ramp up and down and stop more often, collectively termed cycling [10].

Power plant cycling leads to component failures which are complex and usually involve multiyear time lagging. For this reason, a correct estimate of those costs is not an easy task to quantify.

A useful tool in this analysis has been the 2012 Power Plant Cycling Costs report [11] produced by Intertek APTECH for the National Renewable Energy Laboratory (NREL), a National laboratory of the U.S. Department of Energy. This report provides the most recent, available detailed review of real data available on power plant cycling costs that this author could locate.

According to the NREL report, the factors that contribute to the increased cost of CCGTs’ cycling can be summarized as:

  • Increased fuel consumption due to more frequent plant start-ups, lower efficiency at part-load levels and lower efficiency arising from increased wear to components;
  • Increased O&M costs, due to increased wear and tear to plant components;
  • Higher rate of components failure;
  • Increased environmental costs due to higher specific CO2 emissions, due to lower efficiency during start-ups, load following and operation at outputs different from the design one;
  • Loss of income during the plant lifetime, due to more frequent and longer forced outages;

Because of the complexity of the topic and the present lack of other, non-fuel, statistical data, which most generators regard as confidential, the actual costs of more frequent outages are not included in this report. Instead, the author has to rely on the publicly available estimates made in the NREL Report.

The NREL report provides an estimate of the costs associated with frequent CCGT start-ups. These costs depend on the kind of the start-up, i.e. whether it is cold, warm or hot. Cycling damage increases at an inversely proportionate rate to the initial plant temperature, because of the increased thermal stress on the plant components. According to NREL [11], CCGT start-ups after no more than 5 hours offline can be classified as hot starts, between 5 and 40 hours as warm starts and over 40 hours offline as cold starts. It provides to the general public only lower bound, non-fuel costs associated with CCGT’s starts; those are (in 2012 US$):

35 US$/MW capacity for hot starts;
55 US$/MW capacity for warm starts;
79 US$/MW capacity for cold starts;

These numbers are median estimates. The NREL report provides 25th and a 75th percentile values as well, and recommends the use of the 75th percentile values “if a Unit is judged to have more than typical cycling usage and susceptibility compared to the generation type”, and the 25th value “if a Unit is judged to be less susceptible to cycling costs or low cycling usage compared to other units in the generation type population” [11]. Due to the lack of more detailed data, the median value has been used by this author for the analysis of this work.

It is important to remember that the actual cost is highly dependent on the specific plant technology, years of service, components used and operating conditions. The numbers provided to the public are lower bound estimates, making any analysis based on them conservative.

In order to determine the costs associated with each CCGT in the Irish network, metered generation data from the SEMO portal has been used [8]. The SEMO portal provides 30-minutes resolution data for each of the eight CCGTs registered in 2016 in the SEMO market for Ireland. The years from 2008 to 2015, representing those years when wind power became a significant part of the fuel mix, have been analyzed in this study.

Using the previously mentioned methodology, an estimate of the cycling costs associated with the plant start-up has been performed. The lower bound of these costs for the Irish CCGT fleet (8 units) is in the range of US$ 15-20 millions per year. The data also show a generally increasing trend over recent years, due to the increased wind power installed capacity. The peaks in the years 2010 to 2012 are caused by the
non-availability of the East-West Interconnector, which will be commissioned in September 2012. In addition, Turlough Hill pumped storage station was under maintenance between 2011 and 2012, and the Moyle Interconnector was out of service from October 2011 to February 2012 [17], [18]. The lack of infrastructure to export or absorb wind power fluctuations resulted in more intense cycling of the thermal fleet, and consequent O&M costs.

Figure 10: Conservative estimate of the O&M costs of CCGTs’ start-ups.

The 2010-2012 excursion aside, the reason behind the increasing trend are to be found in the increased number of starts and stops, required by the fleet in order to absorb the increasing share of intermittent wind power output, and with the increasing share of cold starts in the total number of start-ups. The last topic that will be covered in more detail in the next section of this report.

Another cost taken into consideration in this work is the high specific fuel consumption occurring during the plant start-ups and stops. A CCGT is a complex machine, which requires a relatively long time from being offline to reaching the rated power output. During the start-up process, more fuel per MWh of generated energy is consumed than during normal operation.

A cost estimate of fuel consumption during the start-up process was performed by a UK consultant [12]. The dataset provides detailed information about a conventional cold start-up of a 400 MW CCGT commissioned in 2004 which is considered fairly typical for the incumbent generators in Ireland.

Figure 11: Power output for a UK 400 MW CCGT commissioned in 2004. Time scale in minutes.

In this case the start-up process, as recommended by the Original Equipment Manufacturer (OEM), took a total of 103 minutes, from the plant being started to reach its rated power output. The net power output results as the sum of the gas turbine and the steam turbine power outputs, and it follow the trend presented in Figure 11.

During the entire start-up process, the 400 MW plant required a total heat input of 2837.9 GJ. Assuming the cost of fuel being 10 US$/GJ, this results in a plant start-ups cost for fuel of US$ 28,379, or approximately US$ 71 per MW of installed capacity. During the start-up the plant produces some electricity, mainly after the steam turbine is activated, as shown by the cumulative plot of Figure 10. During the first 103 minutes of operation the CCGT generates approximately 253 MWh. Because the commercial value of the electricity generated by a CCGT as it is ramping up to its contracted load is a fraction of its commercial value, the value of the electricity generated during the start-up process has been ignored in this analysis.

Figure 12: Electrical energy generated during the start-up process of the 400 MW UK CCGT [12].

The Irish CCGT fleet was commissioned between the year 2000 and 2014, and each plant is equipped with similar but different technologies, which can cause the start-up curve and the specific fuel consumption being different from the one used in the datasheet. In addition to that, the “basis” datasheet refers to a cold start, while many of the actual start-ups are warm or hot, resulting in less fuel to increase the temperature of the plant components and faster start-up time. Finally, sometimes a CCGT is required to be online but at the rated power, with or without the activation of the steam turbine; therefore, the real-time start-up curves are different from the one on which the calculations have been based.

This behavior can be seen from the plot here below, where various, real-time start-ups are plotted for two registered power outputs at the Tynagh and Great Island CCGTs. All the start-ups occur during November 2015. The power output has been converted to per unit values as a ratio of the plant’s rated power. The time scale is in minutes.

Figure 13: Comparison between the power output in the CCGT model and four examples of real CCGT start-ups during November 2015. Data source: SEMO.

It can be seen how the actual start-up process for gas-fired plant can differ significantly from the OEM’s recommended model. It appears that in Ireland some units are required by the system operator to activate only the gas turbine side of the CCGT in case of low demand, resulting in a lower plant efficiency.

Based on the conservative fuel cost of 10 US$/GJ, an average value of 70 US$/MW cap. for any plant start-up has been assumed. The resulting cycling cost figure, sum of the fuel cost and the O&M start-up costs, results to be in the range of many tens of millions of US$. This is shown in the chart of Figure 14.

Figure 14: Conservative estimate of the costs associated with the CCGT fleet start-ups over the years, inclusive of O&M and additional fuel consumption.

It is important to remember that this is a very conservative figure, since the cycling costs estimates are based on the NREL lower bound estimates [11]. The non-fuel costs associated with the increased ramping have not been evaluated due to the lack of transparent data.

5) Analysis of the CCGT Capacity Factor

From the data gathered from the SEMO database, it can be seen that the capacity factor of the Irish CCGT fleet has been steadily decreasing over the last years, from around 80% in 2008 to approximately 40% in 2015. The trend is illustrated in the following plot.

Figure 15: Irish CCGT capacity factor over the years. Source: SEMO.

The decrease in the capacity factor results in an increased share of cold starts in the total number of a CCGT’s start-ups, as mentioned in the previous section of this report, and consequently higher cycling costs. As an example, the number of start-ups of Tynagh CCGT, classified as cold, warm and hot, is plotted in the figure below. As described in the previous section of this report, based on the analysis done in the NREL report [11], cold starts are defined as start-ups after more than 40 hours being offline, warm from 5 to 40 and hot starts after no more than 5 hours offline.

Figure 16: Number of start-ups per type of Tynagh CCGT. Sources: SEMO, NREL.

A similar trend can be seen from the following plot, showing the number of start-ups of the entire CCGT fleet installed in Ireland. As for the cycling costs trend plotted in Figure 10 and Figure 14, the unusually high values for the years 2010 to 2012 were caused by the unavailability of the East-West Interconnector and temporary but extended faults in the Moyle Interconnector and Turlough Hill pumped hydro station.

Figure 17: Number of start-ups per kind of the Irish CCGT fleet. Sources: SEMO, NREL.

The decrease of the CCGT capacity factor is directly related to the increase in wind power installed capacity, and consequently with the increasing share of stochastic renewable energy in the Irish fuel mix. A direct comparison between the wind power installed capacity in the island and the average capacity factor of the CCGT fleet are shown in Figure 18.

Figure 18: Comparison between the average capacity factor of the Irish CCGT fleet and the cumulative wind power installed capacity in the Republic of Ireland.

If the 2020 goal for wind power is met, requiring approximately 4000 MW of installed wind capacity in the ROI alone [13], a linear approximation results in the CCGT fleet capacity factor falling to less than 20%.

The individual CCGTs in the fleet have a widely diversified operation, as it can be seen in Figure 19. It is worth noticing how the newer installations, such as Great Island CCGT and Aghada CCGT (commissioned in 2014 and 2010 respectively), have been used less than older, less efficient units such as Dublin Bay (commissioned in 2002) or Coolkeeragh CCGT (commissioned in 2005).

Figure 19: Capacity factor of the individual CCGT units in the Irish fleet over the years. Source: SEMO.

It seems that location, and therefore possibly the grid’s existing configuration rather than the nameplate efficiency of the power plants, plays the most important role in determining the use of each individual CCGT.

As an example, the capacity factors of Dublin Bay CCGT, located in the Dublin area, and Tynagh CCGT, located in County Galway, are compared in the plot below. The older, less efficient Dublin Bay power plant is operated at nearly the maximum of its potential, probably due to its strategic location.

Figure 20: Comparison between the capacity factor of Dublin Bay and Tynagh CCGTs with the Irish fleet’s average.

During the last few years, three CCGTs – Dublin Bay in the Dublin area, Whitegate in the Cork area and Coolkeeragh in Northern Ireland – provided most base-load operation, while the remaining plants in the CCGT fleet were activated during peak load demand and/or during periods with low or null wind power generation.

6) Proposed Commercial Measures for Mitigating the High Costs of Wind Integration in Ireland

Incoteco (Denmark) ApS is working with two possible technologies that might ease wind power integration in the Irish scenario while lowering its costs. These are MW scaled battery modules and the Trent-Ormat combined cycle gas turbine respectively.
Battery Storage Technology

The Eos Aurora 1000|4000 is a promising candidate for battery storage solutions in Ireland. It can deliver 1MW of energy for 4 hours continuously with a millisecond response. This would allow multiple modular units to absorb rapidly wind power fluctuations and to compensate for forecast errors. Due to its rapid response time, the unit can also provide frequency regulation, an important ancillary service in an electricity island such as Ireland.

The Eos Aurora’s zinc hybrid cathode (ZnythTM) sealed battery technology allows the unit to have a long cycle life (5000 cycles at 100% DOD) and is (to our knowledge) the lowest priced energy storage produce on the global market today. According to Eos’ datasheet, the retail price for DC modules over 10 MW is 160 US$/kWh (640 US$/kW), ex-factory, USA.

The unit has a high cycle efficiency, over 75% at 100% DOD, which is about the same as most modern pumped hydro installations. As a comparison, the only installed storage system now installed in Ireland is the Turlough Hill pumped storage station, which appears to operate with an average efficiency of 55% during 2015 [9]. The most important technical features of the Eos Aurora technology are summarized in the following table:

Table 2: Eos Aurora 1000|4000 Technical Specifications

Battery storage could be integrated into the Irish grid for three main purposes:

1) To reduce the cycling stress on the present thermal fleet, allowing these to be more fuel efficient;
2) To displace the least efficient CCGT units used for balancing and infill power in the fleet;
3) To reduce wind power curtailment, while providing the foregoing services;

Reduce Stress of CCGT Units

The use of the CCGT fleet to absorb wind power fluctuation has proven to be expensive and inefficient in terms of O&M, fuel consumption and CO2 emissions. The wide spread use of such storage systems allows a power system to decouple (to some extent) demand and production, by storing part of the excess energy in times of overproduction and injecting it into the network when needed. Storage capacity in Ireland can relieve the thermal fleet from absorbing part of wind power intermittency; by doing so, cycling, ramping and the high costs associated with these would be reduced.

Battery storage presents significant advantages over the competitive technologies in Ireland. A new, 650 M€ pumped hydro installation, rated at 360 MW, has been proposed for the Irish system [14]. This shows a clear interest from the investors in a major capital investment in energy storage; however, the time scale of the investment does not match with the ongoing increase in wind power installations and the ambitious 2020 targets – only 4 years from the time this report was produced. Modular MW scale battery storage can also be installed before 2020, in many locations, including the built environment and at wind farms.

MW scale battery storage could ease wind power integration into the network, while allowing the existing CCGT fleet to operate more efficiently and with reduced cycling costs.

Displace Least Efficient Units in the Fleet

Battery storage can be used to displace the least efficient operations of the CCGT fleet and help the remaining units in the fleet operate in more optimal conditions. Three CCGTs (Tynagh, Huntstown and Poolbeg) showed a drastic reduction in their capacity factor over the recent years. It is possible that some of these units should be decommissioned for lack of operation in the near future.

It is also likely that the coal-fired units may be closed to reduce carbon emissions between 2020 and 2025, requiring their capacity to be replaced, MW for MW [15].
Partly replacing these thermal units with battery storage could be a solution to improve the performances of the remaining fleet. The installation of MW scale battery storage can efficiently absorb wind power fluctuations and allow the remaining CCGT fleet to operate at higher capacity factors and closer to their design power ratings, therefore more efficiently. The use of fast-acting battery storage to absorb most of the wind power fluctuations would eventually allow the rest of the CCGT fleet to operate with less ramping and possibly fewer start-ups, thus reducing O&M costs, specific fuel consumption and lower specific CO2 emissions.

Reduce Wind Power Dispatch-down

Dispatch-down of wind energy refers to the amount of wind energy that is theoretically available, but cannot be used. Dispatch-down due to overall power system limitations is referred to as curtailment, while dispatch-down due to a local network limitation is referred to as a constraint [16].

During 2014, according to the EirGrid report [16], 4.1 % of the total wind energy generated in the whole island of Ireland was lost due to down-regulation, accounting for 277 GWh. Lost wind power production is currently on an upward trend. The increase in the wind power dispatch-down is directly proportional to the wind power installed capacity and the increase in wind generation connected to the system [16]. A tentative forecast based on a linear approximation is shown in Figure 21.

Figure 21: Wind power down regulation over the last years and linear forecast to 2020. Source: EirGrid.

With the increasing trend of wind power installations, it is unlikely that unremunerated wind power curtailment can increase if no countermeasures are applied. Assuming that Ireland would reach the 2020 target for 40% renewables in the fuel mix, with approximately 13 TWh of energy produced from wind power, a linear approximation of the curtailment figure for the years currently available leads to a down-regulation level around 8% – i.e. 1 TWh. This is clearly only a forecast, based on the available data during April 2016.

The release of the 2015 EirGrid curtailment report, expected in the first half of 2016 and not available at the time this report has been released, would make any analysis on this topic more accurate.

Modular battery storage solutions could be installed directly at the wind farm site, minimizing line losses and contributing towards the easing of the network’s limitations. If placed on the DC side of a wind farm – where possible – this technology has the advantage of no need for additional DC/AC conversion than the one already available on the wind farm site, resulting in reduced installation costs.
However, EOS’s production will not start until the third quarter of 2016 and large-scale production will not start until 2017, for a global market. Where ever EOS batteries are installed, it is likely that the owner operators will require at least three years of proving experience before any mass roll-out at the scale of hundreds of MW can happen before (say) 2020. Even then, global demand at the scale of GW per year is unlikely before the mid-2020s.

Until then, Ireland will need to rely mostly upon thermal power stations to mitigate the growing operational costs and increasing unreliability of its frame-type CCGTs.

Trent-Ormat Technology

This novel CCGT technology is immediately available. The system module is 132 MW, and consists of two 55 MW Rolls-Royce Trent aero derivative gas turbines and a 22 MW (net) Ormat Organic Rankine Cycle (ORC) turbine. Both these components are designed for flexibility; the Trent-Ormat CCGT can start up fast and reach an HHV efficiency of 46% at full load. The gas turbines synchronize with the grid in 5 minutes from a cold start and runs to full load in another 5. The ORC Turbine takes 10 minutes to run up to full load from cold but overlaps with the second 5 minutes of the gas turbine, so the whole plant is up to full load in 15 minutes from cold. This time is reduced to 10 minutes if the ORC turbine is still warm from running within 40 hours.

This technology is well suited for Ireland, where great flexibility is required from the thermal units in order to absorb rapid wind power fluctuations.
This smaller combined cycle gas turbine, in multiple-unit locations, including existing power station sites can supply rapid capacity and allow the incumbent CCGT fleet to perform more efficiently and more often in base-load operation, while the Trent-Ormat installed capacity would be used to compensate wind power intermittent output.

To achieve Ireland’s ambitious targets for CO2 reduction, it will need to close its out-of-IED-compliance coal-fired units by or before 2021. The Trent-Ormat CCGT would ideally replace these MW for MW to maintain Ireland’s need for reliable dispatchable capacity.

7) Conclusions

  • Ireland is expecting to increase its already high share of wind power in its fuel mix during the next years.
  • By 2020, 40% of the electricity production is predicted to come from renewable energies, mainly onshore wind power.
  • Because of the scarcity of interconnections and storage units, the installed thermal fleet – mostly CCGTs originally designed for base-load operation – has to absorb wind power fluctuations. With increased wind penetration, the overall capacity factor of all CCGT’s has halved over the past 8 years from 80% to 40%. Three plants (Aghada, Tynagh and Poolbeg CCGTs) are already operating below a 25% capacity factor. All plants are subject to an increase in power output fluctuations compared with the past. This is causing the gas-fired CCGTs to be operated far below their nameplate efficiency, with consequent higher specific fuel consumption and specific CO2 emissions, therefore of costs to Irish consumers and the unprofitability of investors.
  • New units in the fleet are affected as well, the key factor in determining the use of a power station in Ireland appears to be its location rather than its nameplate efficiency.
  • Apart from the fuel consumption, cycling has high costs associated with the wear of the plant components occurring with more frequent start-ups and stops and ramping. A lower bound estimate of these costs is in the range of millions of € per year for the Irish fleet.
  • Battery storage can be rapidly installed in the island in order to absorb wind power fluctuations, reducing the cycling stress on the gas-fired units. If installed directly at the wind farm site, battery storage can reduce wind power curtailment, an issue expected to become more and more important as the wind power installed capacity increases. However, these will first have to be trialed at different locations and for different applications at as many locations as can be justified commercially.
    The decommissioning of Moneypoint and Kilroot coal-fired power plants will open up a new demand for more flexible gas-fired capacity. The market needs flexible, fast responding and reliably fuel efficient capacity to replace the decommissioned coal capacity and to improve the overall efficiency of the system and thereby reduce CO2 emissions.
  • The Trent-Ormat gas turbine looks like an ideal replacement for coal-fired stations and decommissioning frame-type CCGT, because of its high efficiency and operational flexibility.
  • Investment criteria for new power stations and/or storage units in Ireland are not clear, because of the continuing uncertainty about the possible streams of revenue that a generating/storage unit will be able to access in the next years.
  • In the light of the high and growing costs of integrating wind power, these commercial uncertainties need to be resolved as the issues of supplying non-wind generation multiply to the detriment of investors and the Irish public alike.


[1] Sustainable Energy Authority of Ireland (December 2014). Energy in Ireland – Key Statistics 2014;
[2] IWEA. Wind Statistics. From;
[3] private communication at EirGrid
[4] Global Wind Energy Council (2015). Annual Market Update 2014;
[5] UK Department of Energy & Climate Change;
[6] EirGrid, All-Island Generation Capacity Statement 2016-2025, 2016;
[7] IWEA, Stormy Week Sees Wind Energy Hit Record Generation, November 2015. Available at;
[8] Single Electricity Market Operator. Available at:
[9] Maria Tsagkaraki (February 2016). All Island Specific CO2 Emissions CCGT Fleet and Wind Penetration 2014-2015. Incoteco;
[10] Niamh Troy, Generator Cycling due to High Penetrations of Wind Power, August 2011;
[11] National Renewable Energy Laboratory, Power Plant Cycling Costs, April 2012;
[13] E.V. Mc Garrigle, J.P. Deane, P.G. Leahy, How much wind energy will be curtailed on the 2020 Irish power system?;
[14] The Irish Times, January 2016. Available at:;
[15] Energy & Natural Resources, Ireland’s Transition to a Low Carbon Energy Future Department of Communications, December 2015;
[16] EirGrid, Annual Renewable Energy Constraint and Curtailment Report 2014, December 2015;
[19] IWEA, Wind Energy and the Electricity Market,
Available at:;
[20] CIA, The World Factbook, Available at:
[21] Department of Enterprise, Trade and Investment, Energy in Northern Ireland 2016, March 2016.

This entry was posted in Energy and tagged , , , , , , , . Bookmark the permalink.

72 Responses to Commercial Measures to Reduce the Cost of Wind Integration in the Island of Ireland

  1. Euan Mearns says:

    Riccardo, I have a question for you about Figure 1. The report describes the problems of integrating a lot of wind power on to Irish grid. And yet Figure 1 shows hardly any wind power at all. Presenting data as TOE, the BP convention, which I believe is correct, is to gross up primary electricity by a factor of 1/0.38 to account for thermal losses that have not been incurred. Have you done this in figure 1?

  2. Euan Mearns says:

    Interesting link on Irish CCGTs with ages and costs:

    Aghada 2010 €360m
    Dublin Bay 2002 €230m
    Great Island 2015 €330m
    Huntstown 1 2002 €170m
    North wall 1982
    Poolbeg 1997 $140m
    Tynagh 2006 €300m (Royal Bank of Scotland)
    Whitegate 2007 €400m

    Total €1.93 billion – a lot of capital to write off?

    • A C Osborn says:

      My thought exactly, why would you throw away and replace at high cost Generators that have 20 to 50 years of life left.
      How old are their Coal Plants?

      You get the impression that no cost is too high when it comes to “green” energy.

  3. How much battery storage would be needed to significantly reduce the cycling in gas-fired units and how much would it cost to buy and install them?

  4. Andy Dawson says:

    I’m curious as to why the choice of the lower range impact for the current fleet, as opposed to median or higher?

    The article is also rather weak in the absence of a discussion of how much storage is required to ameliorate the intermittent effects of excess wind. It’d also be good to see what’s the expected carbon intensity of a regime where 40% of production is from wind.

    • Hugh Sharman says:

      Andy, we set out with the intention of making that evaluation. But soon concluded that there is no solid commercial case existing for fresh investment in any part of the system called “generation” until we know that the regulator, CER, is fully aware of the high and increasing fuel and non-fuel costs of using frame-type CCGTs to balance and stabilize the Irish grid.

      The market rules are set to change next year. The capacity payments to the CCGT owners is set to be cut severely and the money transferred to a new pot called “system services”. At the same time, the generators are being pressed into investing into changed rate-of-change-of frequency (ROCOF) settings for their CCGTs that will allow Eirgrid to operate the system with a much higher fraction of non-synchronous power.

      As far as I can see, the generators are far from happy about this situation.

      I feel any that any investment calculation that does not correspospond with the actualities of the market regulations would be pretty hollow and pointless.

      There is, of course, time for OFGEM, UK market regulator to wake up to the actual likely costs of balancing the UK system when wind penetration in GB reaches 20% by 2020!

      • gweberbv says:


        I understand that the increased number of startups of the CCGT plants increases their costs. But one the other hand, on average these plants are running now for only half as much hours compared to 2008. This should also – to a certain extent – reduce the overall costs of the CCGT fleet. Obviously, they are not consuming gas when switched off. And the ageing of the plants should also correspond to the number of running hours. Probably, you have the numbers available.

        So, at the end of they day are there really increasing fuel and non-fuel costs of the CCGT fleet? Or is just one branch of the total costs expanding with increasing wind penetration, while others are decreasing?

        • Andy Dawson says:

          ” This should also – to a certain extent – reduce the overall costs of the CCGT fleet.”

          but not unit cost…

          • gweberbv says:

            But why should the system operators be concerned with unit costs? When one made the decision to slash the market share of the CCGT plants by 50% within a few years (and even more in the years to come), it is clear that the unit costs of these plants will increase.

            If also the total costs of the CCGT fleet were increasing due to increasing wind penetration, then maybe one needs to think of shutting it down and replacing it by something else that is more cost effective (if such thing exists). One can argue that it would have been better to build more flexible CCGTs or whatever in the first place when it was already clear that these plants will end up as a servant for fluctuating wind production. But the decisions were made and the investments were taken. This money will never come back. The only question that remains is how one can manage the fleet in a way that their running costs are not going through the roof. And to my understanding cutting their running time by 50% should significantly decrease these costs (even if some costs blocks are increasing due to more frequent starts and stops).

          • Andy Dawson says:

            “But why should the system operators be concerned with unit costs? ”

            Perhaps because that’s what defines their breakeven point, and whether they can stay in the market?

            “If also the total costs of the CCGT fleet were increasing due to increasing wind penetration, then maybe one needs to think of shutting it down and replacing it by something else that is more cost effective (if such thing exists).”

            Good luck finding that, at a credible technical scale and reliability.

            “One can argue that it would have been better to build more flexible CCGTs or whatever in the first place”

            I doubt when these plants were built that anyone was envisaging such silly levels of wind penetration

            “clear that these plants will end up as a servant for fluctuating wind production.”

            Actually, you misunderstand the role of a system operator – it should be to optimise the system as a whole, not to be the “servant” of one particular form of generation

          • Greg Kaan says:

            One can argue that it would have been better to build more flexible CCGTs or whatever in the first place when it was already clear that these plants will end up as a servant for fluctuating wind production

            And one could argue that wind and solar farms should not have been built until some means of overcoming their inherent intermittency was overcome. It is already clear that until then, wind and solar do not contribute materially to power generation in industrialized countries.

            Stop ignoring capital costs. They are a limiting factor for any economy and ignoring it is a path to disaster (which you have us heading along)

    • robertok06 says:

      “The article is also rather weak in the absence of a discussion of how much storage is required to ameliorate the intermittent effects of excess wind”

      Hi: modern wind generation has become much cheaper than any form of storage via batteries, and therefore, as explained here…

      “The energetic implications of curtailing versus storing solar- and wind-generated electricity”

      … it is cheaper (and easier, to some extent) to over-install wind turbines rather than looking for battery storage of their excess energy. This does not apply to photovoltaics, which are so expensive that almost any form of storage of the little electricity they produce should be preferred over curtailing.


  5. gweberbv says:

    Wouldn’t it already help a lot for the existing CCGT fleet when the coal and peat plants were shut down (maybe kept in reserve)? This would allow for higher capacity factors of the CCGT plants.

    If more interconnectors are built, it is likely that all investments in the existing fleet will end up as stranded assets.

    • Hugh Sharman says:

      @Gwererby, as regards inter-connectors, start by reading

      As dispatchable capacity is closed down by political consensus (populist writ?) all over Europe, the business of relying on neighbours to supply shortages of power will increasingly resemble that old saying of “making a living by taking in each other’s washing”.

      This will be the case quite soon, starting with France’s and Sweden’s ageing nukes! The bill for an already bankrupt EdF well north of €100 billion!

      An adequate self-sufficiency in gas power plant looks sort of prudent to me! But EU wisdom seems to think otherwise!

      • gweberbv says:


        to my knowledge CCGT plants produce relatively expensive electricity. At the same time, interconnectors typically add only a very small portion of extra cost on the price of electricity that is flowing through.
        So, I would expect that it makes economic sense to use cheap imports when they are available. And to have an extra chance to make money with your own fleet of expensive power plants in the case that your trade partner needs additional power.

        In particular when the alternative – that is presented here – is based on batteries and additonal gas plants (on top of the already underutilized fleet).

        • Greg Kaan says:

          guenter, as with many renewables proponents, you continually ignore capital costs and only consider operating cost – I was tempted to point this out to you with your comment about reduced fuel usage for CCGTs when underutilised.

          Interconnectors are far from cheap – especially if they are undersea or underground. And since they do not generate electricity but only transport it, they add a parasitic cost to the power that is delivered in order to pay back the investment required to build them and cover maintenance.

          As an aside, in Australia, wind farms and PV typically have higher operating costs than coal plants per MWh. So their power is far from “free” even discounting the upfront capital cost.

          • gweberbv says:


            just have a look at one of the French-UK interconnectors:

            About 2 billion of investment costs (for 2 GW) and annual operation costs of 5 million (in money from 2001). In operation for 30 years now and it does not look like falling apart anytime soon.
            Let’s assume a capacity factor of 50%, we end up with 261000 GWh of transported electricity. If I assume that the total costs (financing with unrealistic high interests or whatever) were 3 billions, I find transmission costs of 10 Euro/MWh. And each more year the interconnector is in operation, is further reducing this costs.

            Yes, I guess this is very cheap (of course, an extremely long interconnector liek Icelink would be more expensive – and probably therefore is unlikely to be built). And at the beginning of the 80s this was pioneering work. Today undersea cables are more common.

          • Greg Kaan says:

            Very good. And we have the Tasmanian BassLink situation caused largely by the the cost of the interconnector making arbitrage a primary concern.

            In any case, you still need spare capacity on one side of the interconnector for it to be of use. Interconnectors are crutches, not the foundations for a grid

          • gweberbv says:


            but the technical measures proposed in this article are crutches as well. They are not necessary for providing that ‘the lights stay on’, but to mitigate certain aspects of high wind penetration.

            I am just saying that maybe building the Celtic Interconnector (or a compareable link to a much bigger electricity market than the Irish one) would do the same job for less money. And if one would now start installing batteries or new gas plants providing better flexibility, these devices will become worthless once new interconnectors are available (because they have the higher operating costs).

            Clearly, one cannot rely on a single interconnector for providing power 24/7/365 as one cannot rely on a single power plant to do so. One always needs backup.

          • robertok06 says:


            “Let’s assume a capacity factor of 50%,”

            Wrong!… try again, much lower.

          • Stuart Brown says:

            Roberto – don’t get carried away, GW is talking about the Interconector France Angleterre. Admittedly only 54.9% in 12/13, but 83.84% in 13/14.

      • robertok06 says:

        “The bill for an already bankrupt EdF well north of €100 billion!”

        Hugh: Idon’t know where you’ve taken this statement from, but it is completely wrong.
        EdF, true, is not in the best shape of its life, but EdF, the biggest electric company on the planet, is in a much better health of any other competitor of large size. EdF does not have a 100 billion Euro deficit or anything like that… quote your sources, please…. let’s be serious, at least here on this blog!

  6. Willem Post says:


    “At an ex-factory price of 160 US$/kWh (640 US$/kW), this technology is lower in cost than all its known competitors.”

    That is the cost, FOB factory.

    Add shipping, insurance, installation, a building in which to put it, HVAC to remove the heat, etc., and the TURNKEY cost will be about $300/kWh.

    • wil says:


      The Wheatley study of the Irish grid shows: Wind energy CO2 reduction effectiveness = (CO2 intensity, metric ton/MWh, with wind)/(CO2 intensity with no wind) = (0.279, @ 17% wind)/(0.53, @ no wind) = 0.526, based on ¼-hour, operating data of each generator on the Irish grid, as collected by SEMO.

      If 17% wind energy, ideal world wind energy promoters typically claim a 17% reduction in CO2, i.e., 83% is left over.

      If 17% wind energy, real world performance data of the Irish grid shows a 0.526 x 17% = 8.94% reduction, i.e., 91.06% is left over.

      What applied to the Irish grid would apply to the New England grid as well, unless the balancing is done with hydro, a la Denmark.

      Europe is facing the same problem, but it is stuck with mostly gas turbine balancing, as it does not have nearly enough hydro capacity for balancing.

      Fuel and CO2 Reductions Less Than Claimed: If we assume, at zero wind energy, the gas turbines produce 100 kWh of electricity requiring 100 x 3413/0.5 = 682,600 Btu of gas (at an ASSUMED average efficiency of 0.50), then 682600 x 117/1000000 = 79.864 lb CO2 are emitted.

      According to wind proponents, at 17% wind energy, 83 kWh is produced requiring 83 x 3413/0.50 = 566,558 Btu of gas, which emits 566558 x 117/1000000 = 66.287 lb CO2, for an ideal world emission reduction of 13.577 lb CO2.

      In the real world, the CO2 reduction is 13.577 x 0.526 = 7.144 lb CO2, for a remaining emission of 79.864 – 7.144 = 72.723 lb CO2, which would be emitted by 621,560 Btu of gas; 621560 x (117/1000000) = 72.723 lb CO2.

      To produce 83 kWh with 621,560 Btu of gas, the turbine efficiency would need to be 83 x 3413/621560 = 0.4558, for a turbine efficiency reduction of 100 x (1 – 0.4558/0.50) = 8.85%.

      Below is a summary:

      Ideal World…………………………..Btu…………CO2, lb…….Turbine Efficiency
      No Wind gas generation………..682,600………79.864……………0.5000
      17% Wind gas generation……..566,558……….66.287……………0.5000

      Real World
      17% Wind gas generation……..621,560……….72.723…………..0.4558

      Actually, Ireland’s turbines produce much more than 100 kWh in a year, but whatever they produce is at a reduced efficiency, courtesy of integrating variable wind energy.

      For example, in 2013, natural gas was 2098 ktoe/4382 ktoe = 48% of the energy for electricity generation; see SEIA report. This likely included 2098 – 2098/1.0855 = 171 ktoe for balancing wind energy, which had a CO2 emission of about 171 x 39653 million x 117/million = 791.4 million lb. This was at least 791.4 million lb of CO2 emission reduction that did not take place, because of less efficient operation of the balancing gas turbines.

      The cost of the gas, at $10/million Btu, was about 171 x 39653 million x $10/million = $67.6 million; it is likely there were other costs, such as increased wear and tear. This was at least $67.6 million of gas cost reduction that did not take place, because of less efficient operation of the balancing gas turbines.

      In 2013, the fuel cost of wind energy balancing was 5,872,100,000 kWh of wind energy/$67.6 million = 1.152 c/kWh, which would become greater as more wind turbine systems are added.

      It must be a real downer for the Irish people, after making the investments to build out wind turbine systems and despoiling the visuals of much of their country, to find out the reductions of CO2 emissions and of imported gas costs, at 17% wind energy, are about 52.6% of what was promised*, and, as more wind turbine systems are added, that percentage would decrease even more!!

      *Not included are the embedded CO2 emissions for build-outs of flexible generation adequacy, grid system adequacy, and storage system adequacy to accommodate the variable wind (and solar) energy, plus all or part of their O&M CO2 emissions during their operating lives; in case of storage adequacy, all of O&M CO2 emissions, because high wind and solar energy percentages on the grid could not exist without storage adequacy.

      • gweberbv says:


        Wheatley used data from 2011, right? Fig. 10 in this article indicates that this was a year where the Irish CCGT fleet was operated in a particular inefficient way. Would be interesting to see the same anaysis for 2014 or 2015.

    • Andy Dawson says:

      I just took a quick look on Ebay – I can buy retail (including free delivery and 20% VAT) a deep-cycle 50Ah 12 v Lead-Acid unit capable of storing about 0.6 kWh for £78. I’d reckon that to be about $69/kW ex-factory. That’s before I get economies of scale for making a larger unit.

      I’m not sure how $160 is supposed to look especially good? Especially when, as Willem points out there are going to be major on-costs, overheads etc to actually put it into operation. That could safely be assume to increase by 50-100% for an installed, running unit.

      Let’s also compare it with pumped storage = Coire Glas is £800M (€960M / $1160M) for 600MW/30GWh. That’s $38/kWh or $1900/kW. Quite how Ireland manages to spend almost as much on a smaller installation is surprising…

      In other words, the same picture you ALWAYS get when comparing battery with pumped storage – battery looks good on a pure $/kW measure, but appalling on a $/kWh basis. That’s why no-one in their right mind is seriously proposing use for anything other than frequency stabilisation.

      And let’s be clear – looking at those operating patterns, it’s not frequency stabilisation that the CCGTs are doing – as noted, it’s Turlough Hill that does that.

      What this proposal appears to amount to is to replace (at huge capital cost), a fleet of CCGTs with less efficient new CCGTs using a low-temperature back end, and duplicating Turlough’s stabilisation role with expensive battery technology that makes no real difference to bulk energy storage.

      Remind me exactly how this is supposed to look like a good diea?

  7. Andy Dawson says:

    That rather assumes that whoever’s at the other end of the interconnector has sufficient spare capacity at a time when Ireland’s wind output is low….

    I’m amused by the “magical thinking” that interconnectors cause in renewables enthusiasts. All an interconnector is is a conduit to another region’s power system – it doesn’t in and of itself secure generating capacity.

    That’s particularly the case for the Irish grid; there’s only one credible end-point Irish interconnectors – British mainland. And odd as it may seem, our weather patterns and hence wind and solar output are closely correlated to those of Ireland. If Irish wind output is low, it’s highly probable that so is British; if skies in Ireland are dark (like most of the time), that’ll probably by the case in Britain too.

    Which basically says that the flows on any interconnector capacity are overwhelmingly one-way.

    • gweberbv says:


      the Celtic Interconnetctor that is mentioned above would connect Ireland to France. From there you will get balancing service much cheaper than with domestic gas plants or batteries. Ireland is such small that its needs will below the noise level of up- and downramping on the contient. Of course, interconnectors can fail. Therefore, one has to keep enough capacity available to always fullfil the n-1 rule.

      But it is a different thing to keep an expensive plant available to use it maybe one time a year or to use it every few days.

      • Andy Dawson says:

        To quote Eirgrid’s website

        “We are currently at the feasibility study stage of the Celtic Interconnector, with no decision made on whether or not to proceed to project development.”

        Maybe you could explain how a 600km interconnector is somehow supposed to be a better idea than a 63km one (i.e. Moyle?)

        • gweberbv says:


          the advantage of the link to France comes from the fact that to connect only to UK does not help much (at least as long the UK grid is only loosely connected with the continent).

          • Andy Dawson says:

            That’s Andy, not Alex

            I’m not quite sure why connection of a 4,000 MW grid to a 60-70,000MW grid is notably less helpful than connection to the French Grid?

            If you weren’t aware, the population of the whole of Ireland is about 5 million – versus approaching 70 million for Great Britain?

  8. Matthew Wylie says:

    Is there a reason you don’t include Ballylumford(600MW CCGT in NI)?

    • Hugh Sharman says:

      Thank you Matthew!

      Ouchhh! I don’t know how Ballylumford got missed!! Riccardo has gone off for a well deserved holiday.

      He may have a good reason which we are unlikely to hear about for a month. Bear with us please until then.

      I don’t believe it makes much difference to our conclusions?

    • Owen says:

      250MW of BL was decommissioned in January and the station is due to be demolished thanks to the Combustion Directive.

      Hence why they soon need to import electricity from Ireland or the lights will go out.

  9. Pingback: Commercial Measures to Reduce the Cost of Wind Integration in the Island of Ireland | SMIPP Ltd.

  10. Greg Kaan says:

    The article states that batteries could be placed in the DC side of a wind farm. There is no such portion – the turbines synch to and generate AC. An offshore wind farm may well be connected via a HVDC link but the voltage would be far too high unless a worryingly large number of batteries were wired in series.

    Andy Dawson has it nailed on the cost for meaningful battery storage – that proposal is a fantasy.

  11. climanrecon says:

    Is it really true that Ireland only has “baseload” CCGTs and peakers? Do they not also have some CCGTs designed for “load-following”, at least to cover typical demand variations during a working day?

    Wind enthusiasts like to point out that as demand varies the system should also be able to cope with supply variations, but unfortunately wind supply can be falling drastically at the same time as demand is rising, giving a double whammy for the load-followers.

    • Hugh Sharman says:

      Yes, that is true. That was my point in commissioning Riccardo’s study.

      The CCGTs that were originally designed for a system with predictable demand, when they can be operated optiomally, are now being “thrown all over the place” by the system operator in order to keep the system stable.

      As Maria discovered ( ), the CCGT fleet is already operating with a fuel efficiency in the low 30%s and this will get worse.

      We are frustrated by the lack of up-to-date data on the O&M and lifetime (non-fuel) costs which generators guard jealously on grounds of confidentiality! I found Riccardo’s Figure 9 especially revealing.

      If I were an Irish CCGT generator, I would be deeply alarmed by the idea of losing capacity payments, even if exchanged for a range of system services, because the kit they bought all those years ago is quite unsuited to provide these services fuel efficiently.

      If Euan will allow me, I would like to illustrate the cash costs of this fuel inefficiency for Irish consumers!

    • Greg Kaan says:

      The efficient range and ramp response of the frame CCGTs would have been sufficiently flexible in conjunction to cover cover typical demand variations during a working day aside from small peaks that would be most efficiently covered with a modest set of OCGT generators. That is why the Irish installed the turbines that they did – jokes aside, they weren’t stupid.

      But covering the variability of wind (and solar) is another, much more demanding task, even if the demand is flat and why would anyone plan for chaotic conditions that had never occurred? That would just require overbuild of high ramping, low capacity generation – sounds like another form of power generation, eh?

  12. Rob says:

    Has there been a study of costs of UK CCGT when used at low capacity.

    We keep hearing how cheap wind is now at £82 per MWh but how much cost
    has been dumped on to conventional power stations

    • Hugh Sharman says:

      Rob, I commissioned Maria and Riccardo to research and write these reports (mostly) because GB is a much larger “electricity island” than Ireland, having large ambitions for more wind and like Ireand having limited hydro resources.

      Both islands use frame-type CCGTs as their principle technology for balancing wind and delivering secure power.

      The fuel mix is not that dissimilar.

      So, yes, I also believe that the HHV fuel efficiency of GB’s CCGT fleet will decline from 47% today to around the low 30%s by 2020.

      I am unaware that this near certainty is yet recognized by DECC, Grid nor OFGEM

      • Gordon Hughes says:

        Hugh – As you may know, various people, including me, have made your point about the inevitable decline in the thermal efficiency of CCGTs as the fleet is switched to backup mode during the current decade. The arguments have been published in reports, submissions to Parliamentary committees, etc. The main reaction from both official (DECC) and academic sources (the Energy Research Centre) is strong denial, often accompanied by personal abuse. Ofgem and National Grid have no real role – they do what they are told.

        Even when the effects are a matter of simple engineering or mathematics, any suggestion that there may be significant system costs associated with a shift from gas to renewables is treated as blasphemy. One might infer that such people know deep down that the economics of renewable generation is poor but they are desperate to believe that this will change if only current support can be sustained.

        There is a kind of Gresham’s Law in operation. Bad arguments drive out good ones because of a tendency to exaggerate in order to defend a position. So the debates generate much more heat than light.

        • Hugh Sharman says:


          Thank you.

          That is why I have been careful to ensure throughout the kind and welcome cooperation, if not endorsement, of Eirgrid.

          Our results are purely empirical. Their accuracy depends on an acceptance of the data down-loaded from Eirgrid’s dashboard and SEMO. If this is wrong, so are our results and conclusions.

          We are always open to an independent audit. I should hasten to add that inside Eirgrid, the results we came up with have not evinced any surprise. After all, Eirgrid already knows how and why CCGTs are being dispatched on a daily basis!

          I am therefore still waiting for the reaction from SEAI, CER, OFGEM, Grid, DECC and Imperial College!

          I remain optimistic that our results will be courteously and seriously received. I am absolutely determined not to pick a fight with anyone!

          • There was an SEAI study done some time ago that did project that capacity factors from their gas plants would fall as penetration increased though it was not the core of the work. I will see if I can dig it out.

          • Hugh Sharman says:

            Thank you Mr Shanahan,

            We have been searching through the SEAI Library for a study similar to ours, based upon Eirgrid and SEMO data.

            In fact, the Energy Policy study referenced by Maria, being “Fossil fuel and CO2 emissions savings on a high renewable electricity system – A single year case study for Ireland” (downloadable at states in the Abstract that:

            “The results show renewable generation averted a 26% increase in Fossil fuels (valued at €297 million) and avoided an 18% increase in CO2 emissions (2.85 MtCO2), as compared to the simulated 2012 system without renewable generation. Wind averted 20% increase in Fossil fuel generation and a 14% increase in CO2 emissions (2.33 MtCO2). Each MWh of renewable electricity avoided on average 0.43 tCO2 with wind avoiding 0.46 tCO2/MWh. Additional renewable related balancing requirements had minor impacts on Fossil fuel generation efficiency; CO2 production rates increased by <2%. "

            We have written to the authors requesting the data that underlies a conclusion which is at great odds with our own.

            No fortune so far! So if you can find that SEAI paper and share it with us, we would be most grateful!

          • gweberbv says:


            I had a look in the paper by Clancy et al. They assume that without wind the CCGT capacity factor in 2012 had increased only from 58% to 62%. Could you check, how much of the 5.2 GW of gas plants they state for 2012 are CCGT and how much are OCGT? This would allow one to check if the increased CFs they are assuming for the various generators corresponds to the actual amount of additional energy that is necessary to replace the contribution of wind.

          • Hugh Sharman says:


            Good point. Prior to the market crash of 2008, when planning was already under way for most of the new capacity, annual demand had been increasing at the rate of 5% pa.

            As you recall, few of us actually foresaw how bad a crash could be! The typical real life time of a new power station is usually in excess of 6 years.

            With the surplus of capacity, the system operation ought (I suppose?) to be easier and more fuel efficient.

            By contrast, GB will have a very tight dispatchable capacity situation by 2020, when GB wind penetration will be more than 20% by TWh, it would be only realistic to expect a worse fuel efficiency for the GB CCGT fleet?

          • gweberbv says:


            all data are already available in the text above. One just has to subtract the 460 MW of the CCGT plant that was brought online in 2014.
            The main difference between your empiric approach and the one of the authors of the other paper seems to be that their model assumes, that wind does affect the CCGTs only marginally.

          • Hugh Sharman says:


            Thanks. That is a good point.

            However, our report was more focused on how to do something useful to raise the simply awful fuel efficiency of an excellent, modern CCGT fleet, the purpose of which, within the Irish system has been so altered by having to “play second fiddle” to wind power.

            What is your take on our conclusions?

          • Hugh

            The report is called “Quantifying Ireland’s Fuel and CO2 Savings from Renewables” published about 2012.

            If I remember rightly they had several scenarios from no wind to much wind. The gas capacity stayed the same for all but the higher wind ones produced less from gas thereby lower capacity factors.

            If I remember rightly.

          • gweberbv says:


            my take is that the Irish island should shut down the coal and peat plants for the summer. Summer peak demand can be covered by gas, oil (destillates?) – even at zero wind and with all interconnectors broken. Bring back coal and peat only for these times, where demand is high and wind forecast is low. Or to compensate for outages of interconnectors and gas plants.
            By this measure you open a baseload band of roughly 1.5 GW that can be filled by the CCGTs. (I am now taking the numbers from 2012 as a base.)

          • Owen says:

            The SEAI report is seriously flawed. They replace wind with an ocgt in no wind scenario and assume that ccgt actually run more intermittently in no wind scenario !!

          • @ Owen

            Thanks. I do remember having some major problems with it as well at the time. However I thought it might, just might be useful for Hugh.

            I have not had a chance to revisit it and probably will not do so over the weekend; trying to nominate a pub of the year for Wales takes priority.

          • Hugh Sharman says:

            You chosen the right priority, Donoughshanahan! 😉

  13. Ed says:

    Surely NI already (now) has a battery storage system at Kilroot? i.e. they are actually trying this??

    • Hugh Sharman says:



      I might mention that we, at Incoteco, are also involved in bidding into National Grid’s 200 MW tender for fast-frequency reserve (FFR), currently under way.

      This global market for FFR services is already maturing and expanding rapidly.

      My personal take is that lithiums, the dominant battery technology in use for FFR for the time being, will be squeezed out of this market because they will never make their promised price targets AND be able to make enough profit to keep them in the business.!

      In my opinion, lead and zinc solutions are already becoming highly competitive with lithium and the manufacturers are coming up with really smart improvements to old solutions.

      I am NOT going to enter into any dialogue over internet purchased lead acids at ridiculous prices. A battery energy system (BESS) is a complex product and is not the subject of this posting.

  14. Hugh Sharman says:

    Several readers have rightly taken me to task (privately) for my unreferenced assertion that EdF’s liabilities in respect of nuclear decommissioning could cripple both EdF and the French State.

    Mea culpa!

    It was easy to Google this and I found several thoughtful references, among these at from which, if I remember correctly, was the orginal source for my assertion.

    There is a chart showing the different estimates around the world for the cost of nuclear decommisioning which range from the lowest, about €billion/GW (France!) to the UK’s astronomical €2.7 billion per GW!

    Take your pick!

    My very humble advice is not to buy into EdF! On second thoughts, a punt at the bookies that nuclear decommissioning costs of France will shortly (within 2 years?) cause drastic “restructuring” of EdF could be as rewarding as a punt on Leicester winning the League this year would have been a year ago!

    I promise to modify my deplorable tendency to make unreferenced assertions in these columns, in the future!

    • gweberbv says:

      EDF cannot fail. If revenues do not allow for sustained operation of the infrastructure, the government can (and must!) make sure that the lights stay on, NPP sites get properly decommissioned and cleaned up, etc. As long as no major nuclear accident happens this will for sure be much cheaper than what Germany did to foster PV and wind power.

      The big question mark – in my humble opinion – is what will replace the current NPP fleet during the next decades. If they get a permission for 40 years or maybe 60 years does not matter. If they drop out of business according to their age, a lot of capacity will have to be replaced in a short time. And EPR does no look promising right now.

      • Hugh Sharman says:


        Thank you!

        In “saving” EdF (and AREVA) and their legacy portfolio, the already highly indebted French State risks following Greece into the “naughty corner”!

        I agree with you that the EU Energy Coomissioners, along with all the EU’s “great and good” seem to be completely blind to the risk of so much EU dispatchable generating capacity disappearing much faster than they are calculating today.

        Something here about smelly but “un-noticed” gorillas in the corner?

        Exciting times ahead!


    • Andy Dawson says:


      first, there’s a good reason for the delta between French and UK decommissioning costs – reactor technology.

      France’s fleet is almost entirely light-water reactors – PWRs in this case. The UK fleet is a mix of MAGNOX and AGR gas-cooled designs. THe former are (a) larger unit power, (b) physically much more compact, (c) don’t require the dismantling of a couple of thousand tonnes of graphite moderator.

      Second, don’t overestimate EDF’s decommissioning liabilities, or more strictly how quickly they’ll `start to be incurred.

      The only station currently earmarked for closure is Fessenheim; even that will be very borderline as to whether it happens before the end of Hollande’s term. The Republicans would be likely to keep it running.

      What is a lot more concrete is EDF’s uprate and life extension plan – that’ll replace steam generators and implement some additional post-Fukushima safety and resilience measures. It’ll put at least a decade, and probably two on the life of 40+ units. It also increases unit output by 3-5%.

      EDF may well struggle to fund that exercise, but it’s very, very strongly cash positive. THE incremental generation comes in at fractions of a Eurocent per kWh.

      • Hugh Sharman says:


        Thank you! That sounds very convincing! You obviously know your stuff.

        How could has such an experienced operator made such a monumental horlicks of the EPR? You do not have to answer that question which deserves an article, all to itself, on this blog!

        • Andy Dawson says:

          We’ll know one day….

          My thoughts are ( in no particular order)

          – A well meaning, but ill conceived attempt to exceed Europe’s already excessive safety standards by a “brute force” approach of increased system redundancy. That was partly based on designing to the particularly prescriptive French and German regulatory models which only really recognise redundancy as the means to achieve safety. This is expensive, but Arena-Siemens beloved they could offset that by increasing unit size.

          – a failure to understand the implications of that extra complexity and scale on construction.

          – a failure to integrate best practice developed elsewhere, particularly in construction technique. Almost all the competing designs are more or less built on the basis of pre-assembled “modules”- EPR is far more reliant on fabrication “in-situ”.

          – using inexperienced or “out of practice” subcontractors (problems with the Civil build at Olilkuoto, and the pressure vessel issue with Flammanville)

          – in the case of Olilkuoto, getting into a contractual pissing match with the client!

          It’s worth recalling, though that the Taishan build has gone relatively smoothly, following a pretty classic learning curve.

          I doubt EPR will ever be competitive against inherently simpler designs. AREVA was trying to position the design as a “Rolls-Royce” for markets with demanding regulation. However, they’ve been caught out by the emergence of designs with better safety achieved through simplification and passive systems (which we’ll probably struggle to get through French regulation).

  15. Owen says:

    Good research once again. The increase in hot starts is most interesting.

    Some points to note. The interconnectors to UK dont really alleviate the wind curtailment problem as 80% of the time we are importing from UK. Presumably, it would not be in the UK’s interest to start ramping down their ccgt during peak hours just because Ireland has some spare wind. The East West Interconnector has also lead to reduced ccgt capacity factors. Ireland received a 500m grant for it. Quite what its purpose is is beyond me.

    Secondly, the solutions offered seem to small to replace at 900MW coal plant that is a must run in the Irish system, day in, day out to maintain voltage control in that region.

    Lastly, Ireland has surplus dispatchable capacity which Eirgrid hope to utilize to fill the shortage in Northern Ireland, hence the urgency to build the North South Interconnector. More on this here :

    • Hugh Sharman says:

      Thank you Owen,

      We could certainly have used your deep understanding of the Irish system earlier!

      Are you in touch with Eirgrid and CER who are planning to change the Irish market rules as early as 2017?

      I hope so!

      • Owen says:

        I’ve had TWO meetings with the Dept of Energy, they are firmly wedded to the belief that more wind can only be a good thing and batteries will magically solve all of our problems ! Backed up by SEAI, so its full steam ahead, it’ll be alright on the night is the prevailing attitude.

    • gweberbv says:


      could you explain why exactly this coal plant is necessary? I could understand that is is generating electricity quite cheap as long as coal prices are low. But why is it technically necessary? What can it do that the CCGTs cannot do?

      • Owen says:

        Moneypoint has to be running at all times to maintain voltage control on the western part of the island. There is also a line from it that runs to dublin so the midlands are dependent on it too for same reasons.

        It has to be kept running 24hrs a day. This is written in the grid code (so Im not giving an opinion).

        Now one could replace it with two ccgt, presumably they would do the same thing. There is some possibility it will be “Draxed” and run on biomass at some stage in the future. I asked the green party what they had in mind, wind, solar and batteries. How deluded can one get.

  16. gweberbv says:

    Looking at Figs. 15 and 19, one might conclude that it is not wind that does the most harm to CCGTs capacitiy factors, but the installation of new CCGTs while at the same time demand is (slightly) decreasing (cf. Fig. 2).

    In 2010, Whitegate and Aghada brought 2 x 430 MW on the market. In 2015 Great Island came online with another 460 MW. Thus, about 1/3 of the stated 3.7 GW capacity was introduced since 2010. During the same time, the overall capacity factor decreased by 50%. This means todays CCGT fleet has ‘only’ lost 1/4 of its market share from 2008.

    • Owen says:

      There is significant over capacity, so yes this would also contribute to low capacity factors. We have capacity over twice peak and three times average demand.

      Then we go and build an interconnector to provide yet more generation.

      The low (or zero) capacity credit of wind means no plant can be shut down.

Comments are closed.