CO2 Emissions Variations in CCGTs Used to Balance Wind in Ireland


The island of Ireland functions as a single electricity grid linked to the British mainland by two interconnectors with a combined capacity of 1 GW. The Republic of Ireland in the south has set a goal to have 40% of electricity generated from renewables, mainly onshore wind, by 2020. Variable intermittency will be balanced using frame type combined cycle gas turbines (CCGTs). As the level of wind penetration grows the CCGTs need to work harder ramping up and down to compensate for variable wind. This causes increased wear and tear on the CCGT plant and also significantly reduces the energy efficiency of the CCGTs raising their specific CO2 production. During 2014 and 2015, average wind penetration was 22%, the CCGTs produced 575 Kg of CO2 per MWh and the average fuel efficiency was 32% compared with a design specification of 55%.

Guest post by Maria Tsagkaraki and Riccardo Carollo, Incoteco (Denmark) ApSData Source: EirGrid

1 Introduction

The island of Ireland, divided politically between the Republic of Ireland and Northern Ireland, is an example of an electricity island, having limited interconnections with Great Britain only. Those are HVDC submarine cables, the Moyle Interconnector and the East West Interconnector, with a rated power of 500 MW each. A new interconnection between Ireland and France, the Celtic Interconnector, has been discussed by the TSOs of both countries involved, EirGrid and Rté respectively, but the project is still under development and its installation is unlikely to take place before 2025 (EirGrid, 2014).

Ireland has the ambitious target to increase the use of renewable sources in the electricity sector, in order to decrease the dependence from imported fossil fuels and reduce CO2 emissions. For the year 2020 both the Republic of Ireland and Northern Ireland have set the target for 40% of renewable electricity in the electricity sector (EirGrid, 2014), (SEAI, 2014), almost exclusively (37%) coming from wind power (Deaneb, Leahy, & Garriglea, 2013). In 2015, Ireland generated just over 23% of its electricity demand from renewables, the vast majority of which came from onshore wind power (Burke-Kennedy, 2015). This represents one of the highest shares of wind power in the world and the highest for an electricity island (Burke-Kennedy, 2015).

Because of the limited interconnections, as already mentioned above, the integration of a large fraction of renewable energy sources is expected to bring new challenges for the transmission system operator and the electricity customers.

All modern power grids require demand and supply to be constantly in balance, so that the grid frequency can be held within the limits of 49.9 and 50.1 Hz (Xu, Østergaard, & Togeby, 2011). The complexity of this task increases when higher shares of intermittent power sources, such as wind power, are integrated into the network. The SNSP (System Non-Synchronous Penetration) issue, being addressed by EirGrid, is to ensure system stability during periods of high wind penetration (Dudurych, 2014). The SNSP limit is being imposed by the TSO in order to prevent excess percentages of total generation by any non-synchronous sources at any time (Deaneb, Leahy, & Garriglea, 2013).

To synchronize production and demand, part of the excess generated energy can be accumulated with the use of an energy storage technology and re-injected into the grid when needed. The island of Ireland has almost no electricity storage s, except for the 292 MW Turlough Hill Pumped Storage Generator. We found that this installation has a poor average efficiency (around 54%), when compared to modern pumped hydro installations or contending storage technologies.

From the analysis of the charge/discharge patterns of the last years, (pumping at night, discharge during late afternoons) it appears that this is being used mainly to trade electricity exploiting the day/night price difference in Ireland electricity market, rather than to absorb wind power fluctuations. However, we were told by an EirGrid spokesman that it is being mainly used to supply primary and secondary reserve.

At present, the Irish Combined Cycle Gas Turbine (CCGT) fleet is used to fill in the gaps left by the intermittent operation of wind power plants (Clancy, Gaffney, Deane, Curtis, & Ó Gallachóir, 2015). The CCGT fleet was originally designed to achieve high efficiencies with a nearly constant output. Because of the variable operation of these power plants, the average efficiency of the CCGT fleet is dramatically lower than the nameplate 55% of the individual new power plants. The frequent start up and shut down operations required to match the system demand with high wind power share in the network affects the plant’s efficiency and increases the wear out of the plants components. The efficiency is also affected by operating the plants below 80% of their rated capacity; a critical limit below which the efficiency drops dramatically.

A direct result of this way of operating the CCGT fleet is an increase of the CO2 emissions for each MWh generated by the gas turbines. While the increased wind penetration successfully reduces Ireland’s whole system specific CO2 emissions (kg/MWh), the values of the specific CO2 emissions of the CCGT fleet analyzed in this work are over 500 kg/MWh.

The main aim of this study is to establish how the specific CO2 emissions of Ireland’s fleet of frame-type CCGTs, and therefore the efficiency of the fleet, responds to the increased ramping, starting and stopping required by the fleet to balance wind power. The results throw light on the expected performance of incumbent, frame-type CCGTs in Great Britain (GB) where wind penetration, by TWh, is expected to be greater than 20% by 2020, up from 10% in 2015. GB’s fuel mix is sufficiently similar and its dependence upon CCGTs to balance wind is identical to that of Ireland.

2 Methodology

This study analyzes the specific CO2 emission of the incumbent CCGT fleet as wind penetration varies from month to month in the Irish system. The analyzed years are 2014 and 2015 (23 months, from January 2014 until November 2015).

2.1 Data

The EirGrid database was used for gathering all the relevant data for the analysis ( & EirGrid group has been operating the national high voltage electricity grid in Ireland since 2006. The primary goal of EirGrid group is to deliver reliable and secure supply of electricity in the system (EirGrid). It is consists of a group of organizations (EirGrid, SONI and SEMO) in order to achieve better results.

The SEMO database was also used for further pricing data ( SEMO is a Single Electricity Market Operator for the island of Ireland and runs the whole electricity market.

2.2 SNSP (System Non-Synchronous Penetration)
In order to calculate the SNSP limit the following equation was used (Dudurych, 2014):

The SNSP limit was 50% until October 2015 and afterwards increased to 55%.

2.3 Calculation of CCGT efficiency

In order to calculate the CCGT efficiency for the Irish fleet, the heat of combustion and the CO2 emission factor for natural gas have been used. Natural gas used for power generation is very rarely pure methane (CH4). Since 95% of the Irish gas consumption comes from the UK through the UK-Irish interconnector, the UK heat of combustion values and the official UK CO2 Emission Factor (CEF) have been selected for this work (World Nuclear Association, 2016). The average heat of combustion for natural gas results then to be 39.5 MJ/m3, while the CEF accounts for 51 g/MJ. Based on these values, a CCGT unit with a 55% nameplate efficiency µ would have CO2 emissions E equal to 335 kg/MWh. Based on the measured CO2 emissions, it is easy to calculate the CCGT efficiency by using the following equation:

As a numerical example, 500 kg/MWh average CO2 emissions would indicate a CCGT plant with an average efficiency of (0.55*335)/(500) = 37%.

3 Results

3.1 Main Findings

From the analyzed years of the Irish system, the main and important outcome of this study is that the increased wind penetration can successfully reduce the specific CO2 emissions of the whole system. Figure 3.1 presents the rates of CO2 emissions in Ireland compared to the wind penetration for years 2014 and 2015. It is obvious the sharp decrease of the specific CO2 emissions when the wind output increases.

Figure 3.1 CO2 emission per MWh, whole system, wind penetration 22%, 2014-2015.

As already mentioned, the Irish system relies on CCGTs for the grid stability of the system due to the increased use of wind power. The CCGT technology was originally designed to run with an efficiency of 55% or more, which corresponds to around 335 kg/MWh CO2 emissions. Table 3.1 shows the huge differences in the specific CO2 emissions of the CCGT fleet on a monthly range through the years 2014-2015. The frequent starting and stopping of CCGT plants affects their overall efficiency, resulting in increased CO2 emissions.

Table 3.1 Monthly results of the CCGT fleet’s CO2 emissions and wind penetration, years 2014-2015.

In March 2014 the specific emissions of the fleet reached 693 kg/MWh CO2 emissions with a HHV fleet efficiency of 27%. In April 2014, the CCGT specific emissions of the fleet were even higher, 773 kg/MWh CO2 emissions, which corresponds to a HHV fleet efficiency of only 24%. These extreme values can be explained due to a temporary loss of the most efficient CCGT plants during that period. Aghada CCGT (commissioned in 2010), Huntstown CCGT (commissioned 2007) and Coolkeeragh CCGT (commissioned in 2005) are the most efficient Irish CCGT plants and were offline for the entire month of April, as it can be seen in Figure 3.2.

On the other hand, Dublin Bay (commissioned in 2002) and Whitegate CCGTs have been operated at approximately the same capacity factor during both months, 90% and 60% respectively. In March 2014, Dublin Bay provided 37% of the entire CCGT fleet energy production, while this value rises to 46% for April 2014. Poolbeg and Tynagh CCGTs showed a decrease in the capacity factor from March to April 2014, while the Great Island CCGT still had to be commissioned (commissioned in the middle of April 2014). Figure 3.2 can better explain the extreme values of CCGT CO2 emission fleet in March and April 2014.

Figure 3.2 Capacity Factor of the Irish CCGTs plants, March and April 2014.

Figure 3.3 CCGT Fleet, CO2 Emission and Wind Penetration, for years 2014-2015

For the analyzed years 2014 and 2015, the average CCGT fleet CO2 emission was 575 kg/MWh with an average wind penetration of 22%. This mean that the average HHV fuel efficiency was 32%, as it can be seen from Figure 3.4. In 2015, the average CCGT fleet CO2 emission was 540 kg/MWh and the average wind penetration was 23%, which corresponds to 34% average HHV fuel efficiency. The results for 2014 appeared even higher CCGT CO2 emissions and lower HHV fuel efficiency. The CCGT fleet CO2 emission was 607 kg/MWh, the HHV fuel efficiency was 30% and the average wind penetration was 20%.

The results are better being presented in Figure 3.3. Figure 3.4 illustrates the CCGT HHV fuel efficiency compared to specific CO2 emissions, which appears a surprising low HHV efficiency of CCGT plants when they reach 600 kg/MWh CO2 emissions.

Figure 3.4 CCGT fuel HHV Efficiency vs Specific CO2 Emissions, kg/MWh.

As it can be seen from Figure 3.5 the CCGT fleet output decreases with increased wind penetration, whereas the CCGT fleet specific CO2 emissions rise. On the other hand, with almost no wind power, the CCGT CO2 emission fleet is close to 400 kg/MWh, which corresponds to 46% HHV efficiency.

Figure 3.5 CCGT CO2 Emission Fleet, CCGT and Wind Output, November 2015.

Figure 3.6 illustrates the connection between the specific CCGT CO2 emission fleet and wind penetration. An example month, November 2015, can verify that the specific CCGT CO2 emissions rise as wind penetration increases. The same results appear during all analyzed months of years 2014 – 2015, especially those months with high wind penetration.

Figure 3.6 CCGT fleet, CO2 Emission and Wind Penetration, November 2015

Figure 3.7 (example November 2015) shows that system emissions decline as wind penetration increases. However, the specific emissions of the non-wind generation rise as wind penetration increases, as it can be seen in Figure 3.8.

Figure 3.7 Specific CO2 emission per MWh, whole system demand, t/MWh, wind penetration 30%, November 2015.

Figure 3.8 Specific CO2 emission per MWh of non-wind generation, t/MWh, wind penetration 30%, November 2015.

3.2 Side Findings

The SNSP (System Non-Synchronous Penetration) issue, being addressed by EirGrid, was also analyzed for the years 2014-2015 in order to check the system stability during periods of high wind penetration. The increased SNSP is being achieved due to curtailing wind power and altering the ROCOF settings of rotating generators. An example of SNSP rate is being presented in Figure 3.9. As a reminder, the SNSP was raised to 55% in October.

Figure 3.9 System Non-Synchronous Penetration (SNSP), Wind penetration 30%, November 2015

In Ireland there is one storage system, the Turlough Hill Pumped Storage Generator. It has a capacity of 292 MW and we found that its efficiency is around 54%. Turlough Hill is supplying primary and secondary operating reserve for the Irish system. Figure 3.10 illustrates the pumping and generation of Turlough Hill for years 2014 and 2015, while Figure 3.11 presents an example month (September 2014) for a more clear picture of Turlough Hill’s operation.

Figure 3.10 Turlough Hill Pumped Storage Pumping and Generation, years 2014-2015

Figure 3.11 Turlough Hill Pumped Storage Pumping and Generation, September 2014

4 Conclusion

Ireland has the highest wind penetration of any electricity generation system operated as an electricity island in the world, with a target of 40% by 2020. Ireland uses CCGTs plant for balancing the system. This means frequent, stochastic starts, stops and ramping of CCGTs power plants, which results in increased specific CO2 emissions and wear and tear costs of the plants.
For the analyzed years 2014 and 2015, the average CCGT fleet CO2 emission was 575 kg/MWh with an average wind penetration of 22% and an average HHV fuel efficiency of 32%. In 2015, the average CCGT CO2 emission was 540 kg/MWh (average fuel efficiency 34% HHV) with an average wind penetration of 23%. The CCGT CO2 emission in 2014 was 607 kg/MWh with an average wind penetration of 20% and a HHV fuel efficiency of 30%).

The main conclusion of this study is that wind balancing and infill power generation is far more costly than is generally believed at high wind penetration. The results throw light on the expected performance of incumbent, frame-type CCGTs in Great Britain (GB) where wind penetration, by TWh, is expected to be greater than 20% by 2020, up from 10% in 2015. GB’s fuel mix and weather conditions are sufficiently similar and its dependence upon CCGTs to balance wind is identical to that of Ireland.

The findings of this study should be also relevant to other “high wind” electricity islands that rely mostly on CCGTs for system balancing and provide some guidance for planning alternative and more economic technologies for infill and balancing power. Alternative methods for providing infill and balancing power need to be implemented, including more flexible thermal units and electricity storage. Incoteco is working with these alternative solutions.

A novel, thermally efficient CCGT has been developed by Incoteco and is available commercially with the main characteristics as follows:

47 % HHV (51.2% LHV) efficiency achieved 15 minutes after cold start up

  • Hot starts and stops in 10 minutes (synchronous power in 5 minutes)
  • Robust to multiple starts and stops
  • Low wear and tear costs
  • No water consumption
  • Fully field proven components
  • Can be an operated unmanned
  • Lower capex than a ”flexible” frame-type
  • Two years from Financial Investment Decision to commissioning (Austell, Hanstock, & Sharman, 2015).

Another solution could be low cost, robust “power” battery power storage for fast frequency reserve, which is becoming commercially available. At the end of 2016, low cost robust “electricity storage” battery will be also commercially available in MW quantities (by Q4 2016).

5 Bibliography

Austell, M., Hanstock, D., & Sharman, H. (2015). Trent/ORMAT ORC Concept for Future Capacity Needs within the UK Power System. Progressive Energy, Incoteco.

Burke-Kennedy, E. (2015, December). The Irish Times. From

Clancy, J., Gaffney, F., Deane, J., Curtis, J., & Ó Gallachóir, B. (2015). Fossil fuel and CO2 emissions savings on a high renewable electricity system – A single year case study for Ireland. Energy Policy , 151–164.

Deaneb, J., Leahy, P., & Garriglea, E. M. (2013). How much wind energy will be curtailed on the 2020 Irish power system? Renewable Energy , 544-553.

Dudurych, I. (2014). Dynamic Security of a Synchronous Power System with High Wind Penetration. Ireland: IEEE.

EirGrid Dasboard. (n.d.). smartgriddashboard. From

EirGrid. (n.d.). EirGrid. From

EirGrid. (2014). EirGrid plc Annual Report 2014. Dublin: EirGrid.

SEAI. (2014). Energy in Ireland Key Statistics 2014. SEAI. Dublin: SEAI.

World Nuclear Association. (2016). World Nuclear Association Heat Values for Various Fuels. Retrieved 2016 from

Xu, Z., Østergaard, J., & Togeby, M. (2011). Demand as Frequency Controlled Reserve. I E E E Transactions on Power Systems , 26(3), 1062-1071.

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

106 Responses to CO2 Emissions Variations in CCGTs Used to Balance Wind in Ireland

  1. Peter Lang says:

    I am surprised ther is no reference to Wheatleys’ analysis of emissions avoided by wind generation in Ireland. It is the gold standard for estimating the emissions avoided by wind. His paper has been referenced many times by me and others in comments on previous Energy Matters threads.

    Quantifying CO2 savings from Wind power:

    Quantifying CO2 savings from Wind power:

    Free Pre-submission version: Quantifying CO2 savings from wind power:

    Emissions savings from Wind Power: Australia:

    • Peter Lang says:

      Maria Tsagkaraki and Riccardo Carollo

      I should have started n=my first comment with: thank you for this interesting analysis. This is another valuable contribution to the important debate about the cost of CO2 abatement with wind power generation. It is consistent what many other empirical analyses have found.

      From your analysis are you able to calculate the ‘CO2 abatement effectiveness’ of wind power in Ireland at 23% penetration?

      ‘CO2 abatement effectiveness means’ % reduction in CO2 emissions divided by % electricity supplied by wind turbines. Wheatley states it as “ratio of % CO2 emissions savings to % wind power generation”.

      Here I explain the relevance of ‘CO2 abatement effectiveness’ for policy analysis. In short. “The cost of CO2 emissions abatement with wind turbines is much higher than current estimates”.

      “The cost of CO2 abatement with wind power in Australia in 2020, under the RET (Renewable Energy Target) as currently legislated, is likely to be:

      • 2 to 5 times the carbon price which was rejected by voters at the 2013 election

      • 4 to 8 times the ‘Direct Action’ average price at the first auction

      • 6 to 12 times the current EU ETS price

      • 100 to 200 times the international carbon price futures to 2020”

      For much more on this refer my submission to the Australian Senate Select Committee Inquiry into Wind Turbines here (Submission No 259), Wheatley’s is No. 348: go to last page, select display 500, scroll to Submission No 259 or search for Peter Lang (and/or Joseph Wheatley).

    • Hugh Sharman says:

      Mea culpa Peter!

      Maria did this excellent study as an intern student for Incoteco, with very little supervision and help from me. I ought to have informed her about Joe’s excellent work but actually forgot. In mitigation, I have been under unusual stress (for an old man like me) this past winter when I have been consumed by moving house alone, hopefully for the last time.

      Also, I wanted a statistically simple and understandable analysis that can be audited, based only upon the very latest, publicly available data and the willing cooperation of system operator Eirgrid to fill in the gaps. So thank you Eirgrid!

      Maria achieved all this and has to be congratulated.

      I wanted to shoot the common, widely circulated myth, that wind energy drives up an electricity system’s CO2 emissions. It does no such thing as her work shows.

      However, the price of using wind energy to reduce fossil fuel dependency for island systems like Ireland is far, far higher than claimed by the wind lobby groups and the academics who “prove” otherwise from complicated but I suspect entirely theoretical models which are made invalid by suspect assumptions.

      Her results have shocked even me! I already knew that the specific fuel consumption of the Irish CCGT fleet would be adversely affected by its operation as Ireland’s primary “wind balancer”. But not by this much!

      Riccardo is building on Maria’s results from the SEMO data base and will shortly be publishing these here, with Euan’s permission.

      He is revealing in detail how the individual CCGTs in Ireland are being operated to deliver grid stability.

      For anyone who ever doubted it, we can now be sure that frame-type CCGTs are absolutely the wrong equipment for balancing wind power above (say) 10% wind penetration. GB’s wind penetration will be at 20% in only 4 years time! Yet all but a very few of us sincerely believe that the UK’s incumbent CCGT fleet can deliver this efficiently!


      • Peter Lang says:

        Hugh Sharman,

        Thank you for filling in those details. I had no idea. I congratulate Maria for this analysis and agree with your comment “Maria achieved all this and has to be congratulated.” My apologies to the authors for not saying that in my first comment.

        I hope the authors or you may be able to answer my question in my second comment:

        “From your analysis are you able to calculate the ‘CO2 abatement effectiveness’ of wind power in Ireland at 23% penetration?”

        I believe this is the really important figure we need for policy analysis. I’ve explained why in my submission to the Senate Select Committee on wind turbines and in the posts I linked in the comment.

        We now have figures for effectiveness for ERCOT at 4.5% wind energy penetration, Australia at 4.7% penetration, Ireland at 17% penetration and hopefully your analysis at 23% penetration. The values of % emissions avoided versus wind penetration from these studies are plotted together with Herbert Inhaber’s review and summary of previous peer reviewed papers on empirical analyses where all the data is publicly available (see my submission to the Senate Select Committee on Wind Turbines linked above). The trend suggests CO2 abatement effectiveness of wind power would be around 50% at 20% penetration (of course this varies enormously from one grid to another)

      • Owen says:

        “I wanted to shoot the common, widely circulated myth, that wind energy drives up an electricity system’s CO2 emissions. It does no such thing as her work shows.”

        One would have to look at the bigger picture, not just CCGT. For example, oil consumption seems to have increased 57% for 2015 at the peaker plants. There is a need for more fast acting plant during high penetrations of wind. So some of this increase may well be due to the higher wind levels. The ocgt at Aghada ran for 400 hours more resulting in increased emissions. According to Eirgrid there are now 230MW of demand side units (DSUs) which are basically diesel generators. DSUs are increasing every year. Capacity payments are offered and more large industrial facilities are relying on their own generation due to the lavish incentives.

        So the whole system is altered by wind and increased SNSP levels.

  2. gweberbv says:

    Some readers may find this video illustrative:

    As even with zero wind the CCGT units still have to follow the demand curve, contrasting the average efficiency with the design value is misleading. A better comparision is the real-world performance with near-zero wind compared to the situation with high penetration levels of wind.

    The ramping of the gas plants during winter time could be reduced, if the plants were able to direct their output into heat (that could late be used) instead of electricity. However, as you need to adapt the whole infrastrucutre to cogeneration this is only a long-time solution. But as far as I know, Denmark strongly relies on cogeneration which allows – at least in winter time – to moderate the decrease in overall efficiency when power plants are reducing their output below the optimal working point.

    • Willem Post says:


      “A better comparison is the real-world performance with near-zero wind compared to the situation with high penetration levels of wind.”

      I agree.

      Denmark has much more robust connections with other grids, and its level of wind generation has near-nothing to do with domestic consumption, as much of that generation is at night, when demand is low.

      Plans are to have more electrically heated thermal storage, and charging EVs at night, but that electric use increase is only a small fraction of the new wind production increase.

      Here is a calculation, by Wheatley, based on SEMO data, which are more accurate than EirGrid data:

      Ireland’s Power System: Ireland had an island grid with a minor connection with the UK grid until October 2012. Eirgrid, the operator of the grid, publishes ¼-hour data regarding CO2 emissions, wind energy production, fuel consumption and energy generation. Drs. Udo and Wheatley made several analyses, based on 2012 and earlier Irish grid operations data, that show clear evidence of the effectiveness of CO2 emission reduction decreasing with increasing annual wind energy percentages.

      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 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..Turb. Eff
      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.

      NOTE: Gas turbine plant efficiencies are less at part load outputs. If gas turbines plants have to perform peaking, filling-in and balancing, due to variable, intermittent wind and solar energy on the grid, they generally operate at varying and lower outputs and with more start/stops. Such operation is less efficient than at steady and higher outputs and with fewer start/stops, just as with a car. Operation is unstable below 40%, hence the practical limit is about 50%, which limits the ramping range from 50% to 100%. Here is an example:
      Simple Cycle……….100%…..38%….40%…26%
      Combined Cycle…..100%…..55%….40%…47%

      • Alex says:

        “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.”

        Unless part of the problem is the “lumpiness” of generating capacity. This means that large turbines have to run at below 50% output. In a larger grid, such as New England, or England, they can turn off entirely, leaving others to operate at 80% or more.

    • Hugh Sharman says:

      @ gweberby, thanks! Great video! It would be most interesting to see an update covering the period that Maria’s analysis covers!

      As her chart for November 2015 shows, when there is no to little wind, the Irish CCGT fleet operates with an HHV efficiency of 48%, about the same as in the neighbouring GB system.

      The costs of district heating infrastructure connected to central power stations in Ireland, where capacity factors are eroding so rapidly, would be unconscionable. There may be more sense in using electricity spilled from surplus wind power for heating. But I doubt it!

      • Willem Post says:

        My above calculation, based on Wheatley, made about 1.5 years ago, assumes 0.50 for the gas turbines, pretty close to 0.48 actual.

    • matthew_ says:

      Denmark is reducing their cogeneration capacity because it does not match so well with wind. The problem is that the cogeneration has to be run in winter to supply the heat demand during the same periods when their is wind and overabundance of electricity supply. (
      The district heating systems are being instead converted to use heat pumps, direct electric heating, and heat storage to help absorb the wind energy electricity peaks (electricity is nearly free when the wind blows strongly). MacKay showed that it is more efficient to use heat pumps driven by CCGT’s than to use cogeneration.

      • Owen says:

        Thanks Matthew.

        In Ireland, they want to make use of excess wind at night to charge EVs. But as I have told them you would need equal conventional gen for nights when there was no wind so again you would run into the same “duplicate grid” problem

  3. Alex says:

    This is worrying news. I’ve seen graphs of efficiency versus output for CCGT plants and generally efficiency starts to drop off sharply below 60% of rated output.

    In a large system like the UK, it should be possible to run all CCGT plants at either over 80% of rated output, or at zero. This sort of approach should handle the general slew of demand. As demand drops off in the evening, outpiut may fall from 100% to 80%. In some plants, it will then rise back to 100%, and a few plants are ramped down to zero.

    MacKay in Sustainable Energy without Hot Air showed that the slew rate (Delta-GWs / Hour) from wind, even at high penetration levels, is less than that from normal demand fluctuations (though may go on for longer).

    The fact that Ireland’s CCGT is performing so badly, even when wind penetration is low, implies that they are not managing the resposnse to demand variation very well. It seems this problem is not eased by the ineffiicency of the pumped storage.

    Without pumped storage to help, a grid really needs a battery system that can output power of about one CCGT, and for a few hours. This means that other CCGTs can run for several hours at least at 80-100%, before being switched off for several hours – typically on a daily basis.

    What is the size of the Irish CCGT units? If they’ve gone for large lumps of 500MW, then these aren’t going to be easy to manage at the best of times (no wind), let alone with fluctuating wind.

    I suspect what Ireland needs is a couple of 100MW scale CCGTs and about a 100MW / 400MWh battery system. That is quite expensive for a small island.

    • Willem Post says:


      CCGTs, in peaking, filling-in and ramping mode, are normally operated at 75% so they can ramp to 100% and ramp down to 50%.

      At 40% they start to become unstable, so 50% is deemed a safe, low-output operating point.

      No power plants of any kind are ever ramped to zero.

      • Willem Post says:


        Coal and gas generating units are maintained above 50% of rated, for the same reason.

        If a power plant has multiple generators, say (3) 200 MW, then units are put on line as needed.

        That does not work well for coal plants, as they need, from a cold start, at least 5 – 6 hours to get up a head of steam, etc.

        Gas plants, less complicated, are quicker than coal.

        CCGTs and OCGTs plants are quickest.

        I am sure, the Irish are as clever as others managing their fleet of generating units in an efficient, AND safe manner.

        It is increased wind energy, with its irregular output (phase, frequency and voltage), AND its non-synchronized rotational inertia, that is continuously messing with the rest of the power system, making it run less and less efficient, as the article shows.

      • Alex says:

        All plants are sometimes ramped to zero. It’s called turning them off.

        The point I’m making is that on a large grid, with lots of generators which are small compared to the grid, the desired output can be achieved by running all generators at either between 60% and 100%, or at 0% (off or idling depending on the design).

        – Ireland may not have that option, as the generators may be large compared to the grid. Though as you point out below, I’m sure they anticiparted that.
        – It requires cooperation between generators (I’m going to switch off, you other generators need to ramp up).
        – I say 60%. 50% may be possible, but there is some efficiency loss at this level.

        The fact that Irish CCGT efficiency is so poor, even when the wind is not a factor, implies that there are issues that are not wind related.

        • Willem Post says:


          From about 40% to 0, they are disconnected from the grid.

          You should not be commenting on this site, as you know very little about energy systems.

          • Alex says:

            Willem, you really shouldn’t be commenting on this site as you can’t manage a polite conversation or understand what people are saying.

            If you want to resort to non technical insults, then I suggest you comment at the Daily Mail, or something similar in your own language.

          • Greg Kaan says:

            Alex, your notion of turning CCGTs off while allowing others to run at full (or close to) capacity would only work for a wholly state owned generation corporation as only they can bear the expense of idling costly generators (at taxpayer expense). Privately owned generators that are expected to produce a return on investment cannot operate in this manner – see the article I linked about CCGT closures in Germany, in reply to an earlier post of yours.

            Willem’s criticism of your posts is valid due to your lack of knowledge in how these electricity market works in Ireland. What makes “intuitive” sense often ignores realities, particularly when economics are factored in

          • Alex says:

            Greg, don’t CCGT plants already turn on and off – certainly ion the UK market, where they are the principle means of meeting meeting demand over the day/night cycle (with pumped storage and interconnectors being used from one settlement period to the other)?

            The use increase in intermittent sources increases the magnitude of the problem, but it’s not a new problem. CCGT will still need to be switched off – mainly at night. If we look at a snapshot of UK grid – (I have no idea why the colour has gone, but I’ve selected wind and CCGT only), then with wind being fairly constant over the last 24 hours (boring weather), CCGT has varied from 21GW down to 9.7GW, with a total of 32GW of capacity.

            Now, I’m pretty sure that all plants are not running at between 33% and 66% of capacity. Rather most will be off at night, most will be on during the day, and some will be off line for weeks for planned maintenance.

            This works in a market because once utilisation falls to 60%, some of the gas turbines decide that it’s better to shut down, so the utilisation of the others increases. So why can’t that work in Ireland?

            The main issue with introducing intermittent sources is that the average utilisation of the gas plants falls, so they’re trying to make a return on ever falling outputs. The soultion to this is a capacity market – after all, we consumers want two seperate services: The actual supply of electricity, and its availability, even when we’re not using it. Without a capacity market, gas and coal generators will shut down – in Germany, where CO2 emissions are not a priority, that means CCGTs closing and coal staying open.

          • Greg Kaan says:

            Hugh and Owen’s comments below have answered your question about whether CCGTs are being shut down and spun up as generation blocks – they aren’t.

            Aside from the economic reason that I mentioned in a reply to another one of your posts, another factor I forgot to mention is that ramping up a cold or even idling CCGT cannot be done instantly due to the possibility of water damage in the steam turbine so they must be kept running to cope with the sudden fluctuations that the wind turbines can create.

            As for a capacity market, the UK recently tried that and ended up with a vast installation of diesel generators


        • Owen says:

          Wrong, most Irish CCGT have efficiency of between 54-57% which is pretty good. Poolbeg is older and the exception at 46%.

          On the windiest day of the year, the CCGT were ramped down to just below 50%.

          The coal powered station was ramped down to just below 40% – very inefficient for baseload plant.

          • Willem Post says:

            Are CCGT efficiencies at rated output?

          • Owen says:

            Yep at full load.
            Once they go below 40% load, we’re talkin serious inefficiencies

          • Alex says:

            Not sure what you think is wrong.
            – Irish CCGT is efficient
            – It is being run sub optimally, sometimes below 50% of capacity.

            As others have pointed out, the usual strategy is to turn off the CCGT units one by one, and maintain the others at a higher, more efficient output.

            So why isn’t this happening in Ireland?

          • gweberbv says:


            there can be several reasons for running the plants like that:
            – Rat race between several owners. Nobody wants to shut down his plant.
            – Grid being unable to transport enough electricity from region A to region B. Thus, in both locations plants have to keep running.
            – Poor quality of wind forecasts rendering proper planning ramping impossible.
            – More complicated things like plants producing to fullfill long-term contracts without taking care of wind penetration which need to get ordered (and compensated) by the grid operator to ramp down.

          • Hugh Sharman says:


            You are quoting rated fuel efficiency at close to 100% rated capacity. In reality, this does not happen.

            Riccardo, also an intern here at Incoteco, is following up Maria’s study with a detailed look at the way the Irish CCGT fleet is actually operated, from downloading individual CCGT operations from the SEMO web site!

            You are likely to be as shocked and surprised as I was. Lotsalotsa operation at far below CCGT rated output and a surprising volume at open cycle GT output!

            Watch this space for further details. (Euan permitting, of course!!)

          • Owen says:

            There are operational constraints in the grid. Five large units (mostly CCGT and coal) have to be running online at all times to maintain voltage control and enough sync gen. Two power stations have to run at all times in the Dublin region, so these will be both CCGT. So we can see what happens on windy days – CCGT been ramped down to 25% load in Dublin, completely crazy :


            In a system without wind, the CCGT would follow demand so would run at or within 75% of full load for most of the time. So efficiencies would still be quite high. It’s once you go down to 50% and then 40% the real inefficiencies kick in.

            Yes, you are right, SEMO data is a great resource but its in half hour per MW so you have to double the figures to bring to correct unit.

          • Alex says:

            Thank you Owen. That is useful information and would suggest that part of the issue is due to the small size of the Irish grid. To what extent that would be replicated in the UK would be interesting – five large units is trivial for the UK.

            Smaller locations with high renewables – I’m thinking of Lanazarote for example – tend to run on oil fired generators. Whilst vilified in the UK, these could actually be lower emission than CCGT running out of its standard parameters. (As could OCGT).

            The solution, if they want more wind, is more flexible, grid stabilsing, storage, which either comes from pumped storage or from batteries. Alternatively, a few Trent engine derived OCGTs.

            In reference to your respoonse to Hugh, CCGT is used to follow load, but coal and nuclear tend to be baseload, so CCGT provides 40-50% of the electricity, but almost all the load following. So in the UK, for example, a snapshot over the last 24 hours:

            CCGT has varied from 12.3GW to 18.1GW, out of about 30GW, whilst wind has been fairly static. So the CCGT “fleet” has run between 40% and 60%, but I’m sure (or hope) that no individual plant runs at below 60% for any length of time.

            Note that snapshot I showed is a weekend, with steady wind. During the week the demand swing will be greater, and sometimes the wind varies (!), so UK CCGT fleet can swing from 30% to 90% utilisation over a day.

          • Owen says:

            The graph for CCGT efficiency Vs Load is a curve not a straight line

          • Owen says:

            As for Cork, there is too much generation and the old network is not able to cope. So we’ve built some of the most efficient CCGT in Europe but their output is constrained. This region has very significant levels of wind.

            So you need to factor in all these constraints. Capacity and constraint payments are required to keep all this generation viable.

          • Owen says:


            I would imagine there are some kind of regional constraints in UK for a certain amount of conventional generators to be online, particularly in city areas.

          • nukie says:

            So Ireland needs to adopt several thing to make good use of wind power
            a) introduce synthetic inertia and frequency controlls in wind farms – new and existing ones – as othercountires have done.
            b) build a grid. A “grid” where every city has to run it’s own power station to power the city is not a grid.
            c) build and improve interconnectors – use modern VSC-HVDC, which can provide all kinds of grid services and have black start capabilities, thus reducing must run capacity towards zero.
            I hope these things are already on the way.
            The irish power supply system, as it operates today is really very inefficient.

          • Greg Kaan says:

            Roger’s first post here deserves to be reread for reality checks


          • Owen says:

            We have a grid and they want to build more to facilitate more wind.

            Synthetic inertia cannot replace huge heavy rotors. Sorry but that is pure fantasy land.

            You cant just build power stations in the one place and transport the power around on lines. You have to space out your power stations to maintain the voltage across the lines. This is basic power engineering concept. Unfortunately I have to deal with on a daily basis all sorts of green Utopian visions here in Ireland, stuff that cant be implemented because of something called “the laws of physics” but our politicians dont really let such a silly invention like that get in the way of wrecking one of the best power systems in Europe.

            We have HVDC interconnectors, 2 of them. They provide non synchronous power, like wind farms. We import cheaper energy from UK. We export wind energy at night, mostly given away for free. How long will we be able to depend on the UK for dispatchable generation given the precarious state over there as Euan & others will tell you.

            The situation is a mess not because of engineers but because the Greens took over.

          • Alex says:

            “Synthetic inertia cannot replace huge heavy rotors.”

            Are you sure? At the end of the day, both are expressed at the end of a wire. With some batteries and timing circuits, you could build a “synthetic” inertia system and the other end of the cable wouldn’t know the difference.

            A lot of the market for batteries in the US is for grid reinforcement and balancing. They do that with digital circuitry, but if you really had to (don’t try this at you power station), you could build a DC motor at one end of a rotor and a generator at the other end.

          • Kees van der Pool says:

            Hi Owen,

            you wrote:
            “Synthetic inertia cannot replace huge heavy rotors. Sorry but that is pure fantasy land”
            Definitely fantasy land.

            Below an interesting article on the looming problems regarding inertia and primary frequency control. It also shows the Irish windmills are already running curtailed (de-rated, ‘deloaded’ in the text) to allow them to increase power to assist grid frequency response when hitting a pre-set frequency value.

            I don’t think windmills can ever be a substitute for huge heavy rotors. Thousands of distributed inertias rotating at different speeds, connected by delay lines, trying to agree on a single grid frequency?
            I don’t think the windmill manufacturers would be too happy either by the vastly increased demands on the bearings, gearboxes, blades and inverters.

            I think the CCGTs will be forced to continue running at very reduced efficiencies even if only to supply the huge heavy rotor inertia. At this point in time there is very little choice.



          • Greg Kaan says:

            I don’t think windmills can ever be a substitute for huge heavy rotors. Thousands of distributed inertias rotating at different speeds, connected by delay lines, trying to agree on a single grid frequency?

            Kees, I think that is the basic issue causing the limited periods of 100% wind generation by GdV at El Hierro

          • Kees van der Pool says:

            Hi Greg,

            You wrote:
            “Kees, I think that is the basic issue causing the limited periods of 100% wind generation by GdV
            at El Hierro”

            This is not correct. The windmills on El Hierro don’t participate in either the inertial reserve or in the primary frequency generation and can therefore not cause any problems, even at 100% wind generation.
            The inertia and primary regulation on El Hierro is exclusively supplied by the hydro turbines+generator assemblies, spinning wet or dry, and/or the array of diesel generators.
            In case of 100% wind, the hydro turbines take care of the inertia required and the primary frequency generation exclusively. I sent you a link on how this all works (and works very well) a few days ago, if you missed it, here it is again:


            The ‘limited periods of 100% wind generation’ are 100% determined by the availability of sufficient wind to allow for 100% wind generation, nothing else.
            Unfortunately, there is not enough water in the hydro system to add substantially to the 100% wind periods when the wind slacks off to lesser values and the operating philosophy calls for hydro to fill in and extend the DIESEL-less periods.

            The water issue is being analyzed energetically by Roger elsewhere on ‘Energy Matters’ with lots of comments, graphs, pictures etc. Its hard to miss.


          • Hugh Sharman says:

            Owen, I gather from your Danish-based, (!! ??) excellent and informative Irish energy blog, I think we have much to share. Would you kindly get in touch with me as I am no blogger. My email address is and my phone number is +45 40551760. Tak på forhand!

    • Andy Dawson says:

      On an immediate reaction, there’s something of a discrepancy between the nominal 335g/kWh and general accepted figures for CO2 intensity for of around 475. However, the 575h/kWh figure still represents a substantial performance hit.

      If we assume that this is what’s attainable on a 60:40 gas/wind grid, and the 475 g/kWh I’d attainable on an all-gas grid,, plus take a 10g/g/kWh figure for wind, then the aggregate intensity is 350gkWh./ Had there been no performance hit, the intensity figure should be about 325. It gives wind an effective intensity of about 25g/kWh instead of 10.

    • Owen says:

      ” This means that other CCGTs can run for several hours at least at 80-100%, before being switched off for several hours – typically on a daily basis.”

      It would take many hours to start them up again, very very expensive to do and they would emit more CO2 doing this. For example, Dublin Bay CCGT was started only 5 times last year, Whitegate 32 times. Better to start them “hot” than “cold”.

  4. It doesn't add up... says:

    It might be interesting to consider the role of the interconnectors in maintaining the SNSP within bounds. With Scottish power now dangerously dependent on wind, the Moyle interconnector could add to Scottish demand at a critical time – and is unlikely to help out in wither direction.

  5. Joe Public says:

    “In a large system like the UK, it should be possible to run all CCGT plants at either over 80% of rated output, or at zero.”

    And who pays for the plants to arbitrarily stand idle? The consumers, of course. The same consumers already screwed into paying FiTs, ROCs & Constraint Payments to the subsidy-farm owners adversely contributing the problem in the first place.

    • Joe Public says:

      Above intended as a reply to Alex, 10:10

    • Alex says:

      Yes, the consumer pays.

      However, they already pay for them to stand idle. With a relatively low capital cost and a high operating cost, CCGT is a good choice for technology that is going to be idle much of the time. Where the aim is to stay idle most of the time, then early results from the UK capacity auctions suggests that diesel is an even better choice.

      Where the “idling” is due to intermittent renewables, then this cost should be apportioned to the cost of the renewables, at least for understanding and policy purposes.

      • Greg Kaan says:

        “With a relatively low capital cost and a high operating cost, CCGT”

        I’m sorry Alex, this is not the case

        “Where the “idling” is due to intermittent renewables, then this cost should be apportioned to the cost of the renewables”

        An excellent and sensible suggestion – which would instantly price most (already heavily subsidised) renewable generators out of existence. Which would not necessarily be a bad thing. And the idling costs would need to include the higher maintenance that has been mentioned (but not quantified) in the discussions of CCGT ramping.

        • Owen says:

          In Ireland, the cost of testing and changing the ROCOF is been passed on to the conventional generators. The wind farm companies get off scott free. Wind farms not only get subsidized, they get cross subsidized.

  6. Doug Brodie says:

    “During 2014 and 2015, average wind penetration was 22%, the CCGTs produced 575 Kg of CO2 per MWh”

    According to UK statistics, the 2014 carbon intensity of UK gas plants was 365 tCO2/GWh, see Table 5D in

    Thus the Irish gas plant emissions were 57% higher than UK gas plant emissions. Is the Irish gas plant similar to the UK gas plant? If so, the higher Irish wind penetration seems to be having a severely detrimental effect on emissions.

    • Peter Lang says:

      Doug Brodie,

      It’s not that simple. The GB figures you quote are not the emissions intensity of the grid.

      I would urge those who want to understand the issues and how to do the analyses to read the links in my first post. Perhaps start with the peer reviewed paper (second link) if you have access or willing to pay for it. Alternatively read the pre-submission version. It actual contains a lot more information which had to be removed to reduce length for journal requirements.

      If you want to understand the policy relevance, I’d suggest you read one of my posts or better still read my submission to the Senate Select Committee inquiry (see links in my second comment).

  7. davey says:

    ‘Another solution could be low cost, robust “power” battery power storage for fast frequency reserve, which is becoming commercially available. At the end of 2016, low cost robust “electricity storage” battery will be also commercially available in MW quantities (by Q4 2016).’

    Does anyone have further details

  8. Thinkstoomuch says:

    Thank you all involved and posters on some wonderful thought fodder. In different directions.


  9. Roger Andrews says:

    The main and important outcome of this study is that the increased wind penetration can successfully reduce the specific CO2 emissions of the whole system.

    The main conclusion of this study is that wind balancing and infill power generation is far more costly than is generally believed at high wind penetration.

    So where’s the optimum balance between CO2 reduction and cost. Are we there yet?

    • Willem Post says:


      For Denmark, there is very little CO2/fuel penalty, even at 40% annual wind energy, as Norway merely runs less water through their hydro plants.

      For Ireland, being an island grid, the CO2/fuel penalty is very high, even at 17% annual wind energy. See my above calc. Increasing wind % makes matters worse for the poor Irish.

      For years, Eirgrid was in denial, would hardly discuss it, but Udo’s and Wheatley’s studies could not be denied any longer.

      For other countries, the CO2/fuel penalty is between these two extremes.

      Germany unloads its excess energy onto foreign grids more and more hours of the year, so THEY suffer the CO2/fuel penalty.

      Good thing that energy comes to them at near-zero cost, after Germany has subsidized it at about 15 eurocent /kWh.

      Various other losses are not being counted, such as transmission losses.

      The eventual significant reduction of synchronous rotational inertia will increasingly affect grid stability; another wind-imposed cost for dealing with it.

      Wind energy keeps on taking at many different levels.

      • nukie says:

        Willem, one of the nice thing with big grids is that there is no fuel penalty. In a grid with a demand of 300-600GW, switching on or off a CCGT-Plant with 300 MW is changing supply by 0,1% or less. Also the grid is everywhere strong enough to supply a city without locally running plants. So the CCGT ramp up when demand comes from zero to full load, and then ramp down again from full load to zero. No operation at partial load during longer times.
        Usually there is more than one block at one site, and when the first fires up the others get warmed up for pre operation state by this first block, too, as far as possible with waste heat. (for highrt temperatures with steam).
        E.G in Karlsruhe one of the coal blocks is always running as must run due to district heating of half of the city, and keeps tha gas fierd blocks warm for immediate start.

    • gweberbv says:


      the electricity market in general has a much lower ‘reduced CO2 emission per euro spent’ ratio compared to housing insulation. Thus, one will always be off from the ‘optimum balance’.

    • Peter Lang says:

      Roger Andrews:

      So where’s the optimum balance between CO2 reduction and cost. Are we there yet?

      That’s easy to answer. The optimum balance between CO2 emissions and cost is with:

      – zero weather dependent renewables connected to the grid.

      – an optimal mix of nuclear, hydro (where available) and gas – e.g. like France has been demonstrating successfully and cheaply for at least the past 30 years (average 76% nuclear since 1985).

      – in the future, it will be with virtually all nuclear (with a diverse range of different types of plants)

      ERP report shows that all or mostly new nuclear and no new weather dependent renewables are likely to be the cheapest way for GB to reduce the emissions intensity of the GB electricity grid:

    • Alex says:

      “So where’s the optimum balance between CO2 reduction and cost.”


      If they want to optimise it further, they should have a subsidy program to rip out resistance heaters and replace them with heat pumps.

    • Owen says:

      The big question is at what point do the CCGT emissions rise higher than emissions saved by wind ?

      So if CCGT are running at say 25% load behind say 75% SNSP, are the emissions from CCGT higher than the savings made from high wind penetration.

      • Stuart Brown says:

        Good question, but that is surely the point at which nothing makes sense any more 🙂

        It seems to me that the optimum point is not something determined by science but by one’s mindset. If the object is to emit the minimum amount of CO2, then that’s the optimum regardless of cost. If the object is to save money then we’re with Peter and Alex.

        The only way to find a middle ground is to agree a cost equivalence for CO2. Does the £18/tn carbon floor price allow a calculation? (beyond me…) Not sure it even applies in Ireland.

  10. Brian Turner says:

    Brian Turner. Why is the efficiency of Turlough PHS so low? is it the machinery, friction loss in tunnel, or whatever. The figure of 48% quoted versus a minimum of 70% for an older plant is poor.
    Glad to see recognition that thermal plant as these gas stations are react very poorly to the increased level of fluctuation stress and stain caused by thermal cycling way beyond the design levels.
    The obvious solution would be to build more PHS facilities with the capability of quick start-up with turbine runner enclosed in compressed air. There must be sites available that need investigation and survey to start. Then the issue of efficiency due to loading and loss of efficiency due to wear and tear of thermal cycling can be better managed.
    There is also a solution which needs developing – large flywheels operating synchronously, say around 12 MWh storage with flywheel road transportable still. (Composite construction.) Small footprint positioned where needed.

  11. Owen says:

    Eirgrids figures are based on mathematical model and therefore do not represent real world emissions. As they have told me you cannot rely on them to calculate emissions with any accuracy.

    So I have been using EPA reports and semo which have real world data

    • Peter Lang says:


      You might want to read what Wheatley did to check the reliability of EirGrid’s CO2 emissions intensity figures with those submitted by EPA. EirGrid’s are much better for the analysis of emissions avoided by wind because they are at 15 minute intervals and calculate the emissions based on known fuel quantities and the heat capacity of the plant at the power output it is operating at. The only thing they do not take into account is the hysteresis in the heat rates and CO2 emissions intensity during ramping, and the calculation of emissions from cycling also somewhat inaccurate. However, SEAI’s modelling analysis say the inaccuracy due to cycling is only 1%. (All this is from memory – suggest those interested read the reports if interested).

      • Owen says:


        Analysis carried out by SEAI tends to come with alot of disclaimers. In their original reports they made no account of running CCGT on low loads or cycling. I and a few others were responsible for them having to revise their reports to include inefficiencies in the grid due to high levels of wind but there are still faults with their reports – see page 65 below :

        Also :

        • Peter Lang says:


          I recognise there are problems with the SEAI modelling analysis of the all of Ireland grid for 2012, for example:

          However, that was not the point I responded to in your comment. The point I responded to was this:

          Eirgrids figures are based on mathematical model and therefore do not represent real world emissions. As they have told me you cannot rely on them to calculate emissions with any accuracy.

          This is misleading. I understand all estimates of CO2 emissions from power stations over short time intervals in the EU are from models. But, I understand, the emissions reported by EirGrid per generator per short time period are as accurate as any country’s. The US actually measures CO2, N2O, and other gas emissions in the exhaust stack, but their system has significant issues too.

          EirGrid explains

          The rate of CO2 emissions is calculated in real time by using the generators MW output, the individual heat rate curves for each power station and the calorific values for each type of fuel used. The heat rate curves are used to determine the efficiency at which a generator burns fuel at any given time. The fuel calorific values are then used to calculate the rate of CO₂ emissions for the fuel being burned by the generator.


          To do the empirical analyses of emissions avoided by wind, the energy and CO2 emissions for each generator are needed at intervals of not more than 15 or 30 minute.

          My understanding is that the EPA CO2 emissions data is from inventory studies and is not published for each generator unit at the short time intervals needed for the analysis.

          You haven’t said whether you have studied Wheatley’s analysis. It is of the empirical data, and he has checked the EirGrid CO2 emissions data against the EPA data from inventory analysis. My understanding is that Wheatley’s analysis is the ‘gold standard’ for estimating emissions avoided by wind generation and the data he used is as good as can be obtained anywhere.

          Wheatley’s analysis found that CO2 abatement effectiveness of wind power in Ireland in 2011 was just 53%.

          • Owen says:

            Hi Peter

            Yes I am familiar with Joe Wheatley’s analysis.

            This is not my opinion, it is direct from horses mouth.

            When I queried Eirgrid on their published CO2 emissions figures they said they could not be relied on. I have searched my emails for their response but cannot find it unfortunately.

          • Peter Lang says:


            You said that before and I responded to it. I explained that all estimates of CO2 (other than measurements of flue gas as is required by EPA in the USA) are based on models. All methods have uncertainties. You have not stated what estimates of CO2 per 15 minutes of 30 minutes for each generating unit should be used instead of the EirGrid estimates.

            I suggest, if a better alternative data source was available it would have been provided long ago. It hasn’t. EPA doesn’t have them. EPA’s estimates are from inventory analysis and are for the year (I think).

            My interpretation of what you were told is that the EPA figures are the ground truth for the year; and estimates based on integrating the EirGrid CO2 emissions estimates over long periods should not be relied on. For this you should use the EPA figures.

            However, if I am wrong, can provide a link to the alternative source which has more accurate CO2 emissions per generator unit per 15 minutes or 30 minute? Are you able to provide a link to the publicly available data for the years 2011 and since? If you can do so, perhaps Wheatley might be willing to rerun his analyses with the new data and compare the new estimates of emissions avoided by wind with his published analysis.

  12. Here’s Figure 3.8 again

    The Figure description reads as follows: the specific emissions of the non-wind generation rise as wind penetration increases, as it can be seen in Figure 3.8. But what the Figure actually shows is emissions increasing up to 1,000MW and then going flat. Is such an effect real and predictable, and if so how much difference might it make?

    • gweberbv says:


      there is a penalty for moving away from the working point of optimum efficiency. This explains the increase of emissions per MWh generated when wind output in inscreasing. However, I doubt that there is a regular working condition for such a plant where the efficiency drops by a factor of 3 or 4 (relative the opotimum working point). Thus, this increase cannot go on forwever to the point where 1 MWh generated from the gas plants comes at the cost of XXXXXXXXXXX tons of CO2. It is more likely that a plant will simply shut down at a vertain point of underutilization. And this gives room for other plants to ramp up again. As a consequence, one might expect a more or less stable plateau.

    • Peter Lang says:

      It’s a pity Energy Matters does not allow anyone but Roger and Euan to post charts.

      I’d urge you to read Wheatley’s analyses to answer your questions.

      Projecting from his analyses of EirGrid in 2011 to higher wind energy penetration suggests wind will be about 50% effective at reducing emissions at 20% penetration. It was 53% effective at 17% penetration (i.e. in 2011). At 50% effective, wind would avoid just 50% of the average emissions intensity of the grid, and the abatement cost ($/t CO2 avoided) would be 2x the analyses that assume wind power generation avoids the average emissions intensity of the grid.

      This stuff is really important to understand. It has very significant implications for energy policy and advocacy.

  13. nukie says:

    Well, if I look at euans graphs, the CO2-Emission of CCGT in Ireland at 0Wind is about 440g/kWh. This results in a efficiency of wind free generation of 42%. Which means the CCGT perorm very poor.
    @ Willem, even a unmoveable ligniteplant like JAehnischwalde was improved inbetwees so far that it can ramp down to 20-30%. So a additional result of the provided data of Willem Post is, that a “modern flexible” Irish CCGT plant seems to bel less flexible and not much more efficient than a “old and inflexible” lignite german plant. Which is a unexpected result.

    @Roger Andrews, it could well be that at higher rates CCGT aure switched off, while others remain at the same load level than at lower wind power contribution.
    It would be unreasonable to operate a CCGT plant at medium to low load levels if it perorms so bad as the plants in this analysis, it would be the target to concentrate the power on a small number of plants, even in competition. A CCGT plant can offer power cheaper when ramping up towards 100%, than a competitor operating at 50%. So beside grid requirements or inefficient management, a situation with multiple CCGT plants running at half load is no economical stable solution, it would slip to a situation where some are switched off and some are running at full power by economic pressure.

    • Greg Kaan says:

      nukie, this is very similar to the suggestion made by Alex. It can only work with state owned generators or else the CCGT operators would have to be paid from a pool that was scaled on capacity (with some deduction for closure to allow for reduced costs). The workings of such a scheme would be extremely complex and would require all CCGT operators to bid at the same rate,

      • gweberbv says:

        At an envisaged wind penetration level of 40%, it is pretty clear that the fleet of power plants dedicated to provide the residual demand (demand minus wind power) should be subject of a common management.

        • Willem Post says:


          That fleet of flexible generators would have to be suitable to deal with smoothing 40% wind energy on the grid on average, which means it would be dealing with 80% wind energy during windy periods and near-zero wind energy during wind lulls.

          That would be a very extraordinary fleet, if it is fossil fuel fired.

          • gweberbv says:


            power production from hard coal at 7 p.m. on July 24th, 2015 in Germany: 14.5 GW
            And 12 hours later: 3 GW

          • Greg Kaan says:

            What relevance does a 12 hour ramp have? If this response time was available, the whole issue of this post would not exist.

            We all know that wind can fluctuate from gale force to virtual calm (and vice versa) in minutes and that a region the size of Ireland or the UK can be affected in less than hour. That is the response challenge that needs to be dealt with.

          • gweberbv says:


            the maximum ramp rate for hard coal I found was 4 GW/h. Roughly 25% of the maximum output over several weeks. So, the 9.5 GW difference within 12 hours does not reflect the maximum ramp rate. But it reflects the variation in output that is possible on a daily basis.
            Maybe this is more illustrative:

          • Alex says:

            An area like the UK can go from very windy to still in about 4 hours. CCGT can go from zero to full power in an hour or so.

            Start up times of 30 minutes!

            The UK CCGT fleet is typically ramps at 2GW/hour to follow load.

            @gweberdy: Yes, even coal plants can be turned on and off over 12 hours. However, usually Germany exports its dirty coal electricity during the day when solar is delivering, and imports clean French electricity when the sun isn’t shining.

          • Greg Kaan says:

            Gunter, daily demand variations are not the issue. It is the generation fluctuation from the wind turbines that must be coped with.

            Alex, even 30 minute start up from a hot idle state is not sufficient to deal with a wind dropout unless you were going to fire up the CCGTs as soon as wind began dropping. And then what do you do if it doesn’t drop out but fills back in?

            OCGTs are the only thermal generators with the light up time and ramp rates that are capable of following wind turbine fluctuations but they are inefficient and not made to run continously

          • Alex says:


            I don’t think the demand slew is any less than a wind slew.
            – UK demand can increase by a third in 4 hours – typically on a weekday morning. The CCGT fleet has to cope with that. (Say 4GW/hour, or 12% of CCGT capacity per hour)
            – That slew is similar to a “bad case” scenario for wind, which could go from 100% to 10% in 12 hours. (also about 4GW/hr)

            There are of course even higher slew rates – half time at WC final is the usual example, but these are of the order of 1GW over 5 minutes – something that pumped storage and maybe OCGT, and in the future batteries, cater for.

            So the UK has a lot of CCGT that is running during the day, and not running at night. Having CCGT that runs when it’s quiet and not when it’s windy is another head-ache, but no more than demand variation.

            There will of course be “double trouble”, if the wind is falling fast and the demand is rising fast, then you could get a CCGT demand slew of 8GW/hr. But given a high wind scenario will beed a lot more CCGT capacity (or storage, which can respond faster), the slew rate is still only 12% of CCGT capacity per hour.

          • gweberbv says:


            at least for German data I can say that the regular drop in demand from 9 p. m. to 3 a. m. (that is followed mostly by ramping down hard coal) is faster than almost all wind fluctuations on a national level. Germany is bigger than ireland, of course. But as most wind capacity is located in northern Germany the difference should not be too much.

            IF (I don’t know if this is the case) the Irish gas plants are unable to ramp from basicly zero to nearly full output with a few hours and then back down again, THEN this is simply the wrong technology for backing up wind power.

          • Owen says:

            Irish CCGT fleet are often “constrained on” which means they get paid to step in outside of their forecasted output for the day. Likewise, they suffer penalty if “constrained off” where the wind is higher than forecast. All this is available on semo website.

            So I assume they run hot most of the time ready to step in or out.

      • Alex says:

        Doesn’t the grid put out tenders for power in 30 minute slots? As the owner of a 300MW CCGT, I might bid to supply 250MW at £39/MWh. My competitor might bid to supply 250MW at £40. He gets to idle – I get to run at close full power. Problem solved.

        As I want to run to my CCGT for 12 hour at a go, I keep bidding at £39 if I’m idle. Once I’m running, I bid at a very low £20/MWh – so I’m guaranteed to get a contract and as it’s a Dutch auction, I get something above £20/MWh.

        Balancing services and Short Term Operating Reserves (STOR) are run under seperate payment schemes.

  14. Davey says:

    I was told that CCGT not the right technology for use with wind power and we should use more gas engine type generators as shown on link

    • Greg Kaan says:

      If you look at the specification of those generators, they are all low capacity (largest is 10MW) and they have an efficiency just under 50%
      Fewer OCGT plants would do the same job.

      The underlying issue is that the CCGT plants are needed because they have the capacity to supply the grid when there is no wind plus they were mostly built prior to the mass deployment of wind turbines (Wexford being the exception), hence the “incumbent” description of the CCGT fleet in the report posted by Euan.

  15. Davey says:

    Has Energy Matters reported on the costs associated with wind and CCGT
    though load balancing & loss in market share

    We have all heard the claims that wind in now the cheapest form of energy

  16. stone100 says:

    In the EPRPUK report they modeled future scenarios where almost all of the electricity was supplied by wind and solar with just the odd few gaps being filled by gas. (I’m meaning something like their “RE.100 +48h store” scenario on page 21 where 12% gas was combined with 88% wind and solar) What sort of efficiency might that gas generation have? How much would have to be kept hot and spinning? Could CCGT do that job at all? Would it need to be done instead by OCGT or natural gas internal combustion engine generators? .

    • Alex says:

      88% wind and 12% gas!

      However, 88% average renewables would be fluctuating between about 5GW and 150GW, and massive storage is needed. They’ve put in 30GW of capacity for 48 hours. That 30GW – wether pumped storage or battery, provides a massive slew capability so any CCGT could be optimised. You would basically be running the CCGT at close to 100% capacity for several days, to top up the reservoirs, and then turning it off.

      However, 30GW x 48 hours is nowhere near enough storage – in terms of energy or power. If demand, including heating and vehicle needs, is 90GW, then storage + gas power would need to be able to provide 85GW.

      And as for energy, given the consequences of a week long black out, then the grid would have to plan for the worst (renewables) weather in 1,000 years, not just a particularly bad January.

  17. Kees van der Pool says:

    Apologies for OT subject. Maintaining grid stability while increasing RE penetration increases capital cost:

  18. Rob says:

    Is it possible to summarize how much loss in efficiency can we expect at high wind penetrations

  19. Owen says:

    Another interesting thing about the Irish system, is that despite having some of the highest amounts of wind penetration, Eirgrid estimate that wind will only meet 3% of peak demand by 2025:

  20. BMD says:

    @ Maria and Riccardo: have you built:
    1- a diagram, by produced kWh, of CO2 emissions of the total Irish production as a function of wind penetration in the grid?
    1-a diagram, by produced kWh, of the price of electricity for Irish households as a function of wind penetration?
    I would be very interested by these items !
    Many thanks for this study!

  21. Hugh Sharman says:

    at BMD, Ummm…yes! Please read the article!

  22. Pingback: AWED Energy & Environmental Newsletter: May 9, 2016 - Master Resource

Comments are closed.