Off Grid Electricity keeps Lights On

Significant amounts of Diesel generation keeping the lights on at times 

 "With regard to demand side units, even at the start of September we saw a significantly tight period. At that point, wind was giving us very low electricity. Demand side units were asked to become active and they did. They can respond and they can make a material difference there " - Energy Regulator, October 2021

Demand side units are off Grid forms of generation that allow large users of electricity to switch off from the grid and generate their own.

According to eirgrid, most of this form of generation is diesel generators :

Industrial generation refers to generation, usually powered by diesel engines, located on industrial or commercial premises, which acts as on-site supply during peak demand and emergency periods [Eirgrid Report]

The energy regulator, in an Oireachtas Committee debate in October stated that these generators made a material difference in September when wind speeds were low.

What this means is that at times a significant proportion of electricity produced in this country is not counted in the official figures. So when you read that gas has produced 57 % and wind 35 %, these figures refer only to the on grid proportion only. We do not have figures for off Grid which is mainly diesel. 

Peak Winter Demand Arrives

Things get a little bit shaky !

 Last night at 5.30pm, the electricity grid hit peak demand for All Ireland at 6,638MW - not far off Record peak demand of 6,878MW reached on December 21st last year.

.

But total system generation was only 6,106MW leaving a shortfall of 532MW.  

Wind energy was low most of the day, only 380MW or about 7% was available for the whole island at 5.30pm.

The two UK interconnectors saved the day with combined imports of 450MW. There was still a shortfall of about 80MW, made up presumably from demand side units. These would comprise mostly of diesel generators and combined heat and power units. These units are "non-centrally monitored" according to Eirgrid and are not included in these graphs.  

Of course, that damned Brexit lot across the sea with their dastardly nuclear power charged us handsomely for the imported power, at € 2,000 a MW. 

To give some credit to the Irish grid operators, they were correct to build the East West interconnector as it is making up for the once efficient gas plant that have been prematurely wrecked from backing up the wind. One just hopes that the UK will have sufficient power to give us on those cold winter nights over the next few months. 

Precarious Winter Outlook

According to the Eirgrid Winter Outlook, the Irish electricity system will be operating at twice the level of acceptable risk this winter. The system is expected to enter the Alert State at times of low wind, low interconnector imports and low temperatures. No mention is made of Huntstown or Whitegate power stations. Are they assumed to return as expected? We are not told [update: it appears that Huntstown will be back later this week]. 

There are a number of engineering realities that are laid bare in the report that are a sobering read and at odds with the endless spin that has been published in the past in the media about green energy. 

•  Only 9% of total wind energy capacity is deemed as reliable or can be relied upon. 

•  Forced outage rates (the rate at which power stations are breaking down) have increased by 5 times over the past 5 years. This would appear to indicate that these power stations cannot cope with higher levels of intermittent wind energy.

•  Some CCGT (gas) plant is scheduled to be unavailable for 5-6 weeks of the winter period because of scheduled maintenance. This will be in November and early March. There is no guarantee that wind energy will be available during these weeks and, hence, the highest risk of blackouts will occur during these periods. This proves that wind energy cannot replace, nor is it equivalent to, a power station. 

There is also another factor which does not seem to be included in this report. Wind farms, like power stations, also need maintenance - in particular, the older fleet. Whilst it is unlikely that a significant amount of them would break down at the same time, sourcing replacement parts may take more time than normal in the present supply chain crisis. 

This is the first time that Eirgrid have published a winter outlook that warns of a high risk of blackouts. Last year, they were concerned about a trend of "increasing demand, dispatchable generation exiting the market and increasing generator forced outage rates". 

They were correct, that trend has continued into this year. Did anyone listen or take note ? Of course not. But then Eirgrid's own chief, Mark Foley, dismissed concerns about blackouts, saying  people can sleep soundly in their beds this winter.  

We must trust the plan. 

Coal and Oil to the Rescue During Cold Weather

Last week was one of the coldest spells of weather here in Ireland of recent years. As usual with very cold periods, wind generation was low. The grid operators struggled to keep the lights on and many amber alerts were issued.

Wind energy contributed about 20% of the power on average. 

Whitegate Gas power station was and still is out of action which is surprising for a modern ten year old power station. Gas power still provided the majority of the power in the grid mix -  around 50%. 

Wind energy last week - only a small portion of the total installed wind capacity of 4,000MW was available at times

 

Coal provided 12% despite only two out of the three generators at Moneypoint functioning.

An internal ESB memo shows that national grid operator, Eirgrid, asked the company to start up one of its three generators at Moneypoint in Co Clare, which had been shut down, to avoid any possible risk of blackouts [Eirgrid].

Starting up a coal generator takes at least 16 hours to start up from scratch (cold start) so presumably this unit was kept ticking over (warm start). This is what many wind advocates do not understand - you simply cannot switch off a large power station and expect to turn it back on again at short notice. 

Imports were only 1% presumably because UK had no surplus electricity of their own.

This then leaves "Other" at 10%. 

There are only two possibilities for what this comprises now that all but one of the three peat power stations have been discontinued - waste to energy and oil/diesel. Only one waste to energy plant is currently in operation at 62MW. So assuming it was running at max output it was providing about 1.5% of the total fuel mix. Edenderry peat power station now operates at about 60MW also (the other half of it's fuel source is biomass) so likewise about 1.5%. Therefore, unless I'm missing something, about 7% of the fuel mix came from oil and diesel generators. 

Which is roughly the portion of fuel mix from the two peat power stations that were closed down. Oil generation has not contributed this much since the 2000s. This amounts to an indictment of the renewable energy program, in that 4,000MW of new wind energy installed cannot replace 230MW of peat.

Based on that, the expectation that Moneypoint coal power station will close down by 2025 is now looking very unlikely.

Only Ten Percent of Wind Capacity can be Relied Upon Over Winter

Eirgrid, the grid operator, have assessed that only 10% of the entire wind generation fleet in Ireland can be relied upon during the high electricity demand winter period. Solar power fares even worse as zero solar capacity can be relied on as the dark cold evenings draw in.

Conventional gas and coal generators are considered to be around 90% reliable as they can be switched on and off when called upon, once sufficient notice has been given. However, wind is unreliable in nature and is therefore given a "capacity credit" to take account of this unreliability :

 Given the variable nature of wind power, the wind capacity credit expresses how much conventional power generation can be avoided or replaced by a certain level of wind power. 

Only 476MW out of a total of 4,617MW or about 10% can be relied upon in the event that a conventional generator fails :

When it comes to keeping the lights on there is no room for the hyperbole and feel good talk that permeates much of the debate around renewables in the media and in the chambers of the Irish parliament.

The spare generating capacity above electricity demand is called the capacity margin. It has halved from 3,199MW in 2016 to 1,793MW in 2018. Two gas powered generators have closed down in 2018 and Moneypoint coal power station has suffered an outage. 

Generators shut down	Output MW
Aghada AD1 (gas fired steam turbine)	258MW
Marina (open cycle gas turbine)	95MW
Moneypoint (coal)	855MW
	1,208MW

In recent days, it was gas power, the UK interconnector and "other" generators (presumably peat and oil) that ensured the lights stayed on in the absence of Moneypoint and the above units. Renewables (mostly wind energy) varied between 7% and 27% so could not be depended on. Back up power stations are still essential in a grid with large amounts of renewables.

You can also see the importance of the interconnector with the UK (EWIC). Can the UK guarantee they will have spare capacity to export electricity to Ireland ? The black line below shows the profile for coal power in the UK in recent days. As you can see, they too were having problems with their coal units.

Scotland too is in a precarious position and will likely be reliant on imports from the UK over the winter period :

At the time of writing the Torness 2 and Hunterston 3 and 4 reactors were down, leaving Scotland with only 612 MW of nuclear capacity and only 3,900 MW of dispatchable capacity to service over 4,000 MW of demand. However, all three reactors are scheduled to restart by mid-December, and any shortfalls that might occur in the meantime could probably be covered by imports from England [Euan Mearns]

If we get a very cold winter, and there are indications that a cold one is coming, the already strained grid could come under a lot of pressure as demand increases. Couple this with more generator outages, and we could be in for some big problems. This is where the demand side of the equation would become crucial - essentially disconnecting large energy users from the grid who would have to rely on their own generation i.e. diesel generators or combined heat and power. 

Essentially, a return to the pre grid times of the early 1900s, when there was a large and inefficient diesel generator in most big towns in Ireland. 

North South Interconnector Update

In 2017, Eirgrid published a generation adequacy report that showed Northern Ireland in a net deficit position by 2021. 

This was partly the justification for the North South Interconnector :

But now in Eirgrid's most recent report, there is strangely no longer a capacity shortage for Northern Ireland :

It seems that Kilroot power station will be kept open for another few years. But even then, there is still no forecast deficit.  This means that the only justification for the interconnector now is to ensure the iSem (All Island single electricity market) functions as a single market. But as Pat Swords showed, at a cost of € 286 million, the payback for the project would be around 25 years, something that would not stack up in the private sector. 

Eirgrid Increase Spinning Reserves - Irish Energy Blog Vindicated

To maintain a stable grid, Eirgrid have always ensured that there is enough "spinning reserve" (or back up generation) running or available to provide power at short notice in the event that a power station trips (i.e. suddenly has an outage) or there is a sudden change in demand. 

For many years, the minimum for this reserve requirement was set at 440MW, made up of four different types of reserve with different reaction times, of 110MW each. The quickest can deliver in less than five seconds, the slowest in less than five minutes, with the latter capable of lasting much longer than the former. The remaining reserves can react within 15 and 90 seconds. 

The two fastest reacting reserves are called Primary and Secondary Operating Reserve and are provided by units already running on the system that can change their output quickly to deal with unexpected events. The two slower reacting reserves are called Tertiary Operating Reserves and are provided by units both already on the system and that can start quickly at short notice. For additional security, there are also replacement reserves that can start from 20 minutes to four hours and are provided by fast acting offline generators such as open gas cycle turbines (basically jet engines).

The question then arises - if wind energy is inherently variable, just as demand is, or unpredictable just as a generator outage is, what impact does it have on reserve requirements in the event that it unexpectedly rises or falls ? In otherwords, does it contribute to an increase in unexpected events that can't be forecast by the grid operators ? If the answer is yes, then more reserves will be required, likely in the form of fast acting fossil fuel generators (and with a consequence increase in emissions.) Consider that the single largest generator that can fail in the system at any one time is about 500MW compared to combined wind energy capacity of 3,000MW which if acting in unison (as it usually does) is six times the size.

In 2014, the SEAI issued their  "Quantifying Ireland's Fuel and CO2 Savings from Renewables" report based on the contribution from wind energy during 2012. 

This is what they concluded in relation to the potential impact of wind generation on reserve requirements :

Future planned increases in wind capacity will influence the reserve requirements, particularly tertiary reserve requirements. The All-Island grid study showed that additional reserve requirement in hypothetical 2020 scenarios is related to the amount of wind installed but that the largest contributing factor remains the loss of the largest conventional unit. Wind power does not necessarily require larger amounts of primary and secondary reserve, when the characteristics of the wind are taken into account in the calculation of reserve requirements. The relative electrical isolation of the All-Island system means that the reserve levels consider the need for a high degree of generator flexibility, while additional rules ensure a sufficient number of units remain online to ensure frequency and voltage stability. Reserves allow the electricity system to respond to unexpected events but the ability of the system to incorporate variability and uncertainty due to renewable electricity generation is primarily determined by system flexibility. 

At present, renewable electricity generation on the All-Island system does not influence the quantity of reserve required.

In essence, the SEAI are claiming there will be little impact on reserves from wind power but with caveats thrown in about it impacting mainly tertiary reserve and careful use of words like "at present" which would indicate that future levels of wind energy are not being examined in their report anyway. 

In the same year, I wrote an article for this blog challenging the SEAI's report and in particular their omission of the impact of wind energy on reserves. I argued that higher levels of wind energy would indeed lead to an increase in reserves and based my argument on research done by both Danish and UCD researchers (in 2007 and 2005) : 

There should be enough spinning reserves to cover an outage of the largest unit in combination with a fast decrease of the current wind power production. However, the capacity of the largest online unit changes dynamically. (Doherty and O’Malley 2005) further demonstrate the dependency of the demand for TR1 [Reserve Type 1] from the installed wind power capacity. 

"Generally, the demand for replacement reserves increases with increasing wind power capacity installed. 

The occurrence of high demands for replacement reserves is mainly driven by a high number of simultaneous forced outages that happen simultaneously to relatively high wind power or load forecast errors. The value of these peaks tends to increase with increasing wind power capacity installed."  - Wind Variability Management Studies (P.Meibom et al)"

My article has now been vindicated four years later as a recent Eirgrid document shows that they have increased the minimum levels of spinning reserve required for the Irish grid.

Previous Operating Reserve Requirements

New Operating Reserve Requirements, Summer 2018

This means the minimum reserve now is 540MW (135 * 4), up 100MW, with at least half of that coming from units already running on the system. The reason for this increase is due to DS3 System Services Contracts. DS3 services, as explained before on this blog, are services required by conventional and other generators to facilitate high levels of wind energy. 

From the onset, the integration of wind generation presented a range of challenges previously unseen in the power sector. Through collaboration with the Regulatory Authorities and the wider electricity industry, DS3 has developed a number of innovative and progressive solutions. 

The cost of which may well become significant :

This means that some types of service providers could be available and eligible for payments for every hour of the year assuming they are not forced out or scheduled out for maintenance, even if the service is not required from those providers for all of these hours. The scale of overall payments will therefore increasingly depend on the portfolio of service providers and the expected availability of individual service providers.

2014 Irish Energy Blog article : http://irishenergyblog.blogspot.com/2014/12/seais-quantifying-savings-from.html

Overcoming Grid Constraints Fails to Solve Inherent Problems with Wind Energy

In 2014, the maximum level of wind energy allowed into the grid was 50%. In 2017, this was increased to 60% and by November of last year trials were run at 65%.  So some of the obstacles to higher levels of wind energy such as grid constraints have been partly overcome, which theoretically speaking should result in higher wind outputs from individual turbines (the capacity factor). 

In 2014, the capacity factor for wind was 27%. In both 2016 and 2017, the capacity factor remained at 27% despite the higher wind penetrations allowed. 

An analysis of wind speeds shows that wind speeds were fairly similar for those years, with 2015 being somewhat higher.  I took a sample of six weather stations from around Ireland, the average wind speeds I obtained neatly fitted with the capacity factors for wind. 

Year	2014	2015	2016	2017
Average wind speed (knots)	9.6	10.4	9.3	9.7
Capacity Factor Wind	27%	32%	27%	27%
Max wind penetration (SNSP)	50%	55% Trial from Oct	55% Perm from Mar	60% Perm from Mar - 65% trial Nov

As can be seen from the last part of the table above, we went from allowing 50% wind into the grid to 60% and by the end of last year 65%. As wind had more access to the grid, we should have seen a higher capacity factor for wind.

This seems to suggest that we have already reached saturation point for wind energy. I would be interested to hear what people think. I have already written about market cannibalisation and diseconomies of scale. Here is strong evidence that supports that argument. Most of the best sites for onshore wind have been used up. The turbine layout at some sites is too dense and newer larger wind turbine models have failed to deliver any significant additional output. And after all, the wind resource itself is limited, particularly in Midland regions. 

Sources

Eirgrid Constraint Report 2017

http://www.eirgridgroup.com/site-files/library/EirGrid/Annual-Renewable-Constraint-and-Curtailment-Report-2017-V1.pdf

Wind speeds from Met Eireann website (in knots)

https://www.met.ie/climate/available-data/historical-data

	2014	2015	2016	2017
Cavan	6.3	6.8	5.9	6.3
Kerry	9.5	10.3	9.1	9.3
Donegal	14.4	15.3	14.3	15.1
Cork	12.2	13.1	11.8	12
Tipperary	8.3	8.8	7.9	8.2
Carlow	7.3	8.2	7	7.3

All stations record wind speeds at 10m above ground level.

Note that total wind output did increase in 2017 by about 18% by adding an extra 530mw of wind capacity, an increase of about 20% on the previous years installed wind capacity. The capacity factor measures the actual output in relation to potential output if the entire wind turbine fleet had been operating at full output for the entire year. So theoretically if wind speeds increase so should the capacity factor. Or if wind speeds stay the same and the maximum level of wind permitted into the grid increases, then capacity factor should also increase.

Delay in New All Island Electricity Market Raises Questions over Future of Power Stations

The new All Island Electricity Market (I-SEM) was set to come into operation in May this year but has now been delayed by six months. This raises questions over what will happen to two power stations set to close this year. Both Huntstown CCGT in Dublin and Kilroot in Northern Ireland failed to secure capacity payments in the recent I-SEM auctions and signaled their intention to close down by the end of May. This would be disastrous for both Dublin and in particular Northern Ireland, which, with no replacement generation ready, would face the prospect of prolonged blackouts for the next year at least. With all the focus being placed on renewables and battery storage, there still has been no proper impact assessment (nor media coverage), now needed more urgently than ever, on how the Irish grid will cope without these power stations.

The blame for the delay in the I-SEM has been put down to software problems. Which is something this blog highlighted a few years ago - the increased administration costs and problems that could arise from the increased complexity of operating a grid designed around wind energy. 

Storms Linked to Power Station Trips

Eirgrid published this list of power station trips from the past few months:




I couldnt help but notice that most of these dates coincided with storms or very windy conditions:

5th october - Cyclone Xavier
12th october - Hurricane Ophelia
16th october - Hurricane Ophelia
21st october - Storm Brian
27th November - gale force winds
24th December - record wind penetration on the grid
3rd January - Storm Eleanor
17th January - Storm Fionn

It appears that as very high amounts of wind generation is allowed into the grid, the frequency can drop to a dangerously low level.

One of our key tasks is to maintain balance between electricity supply and electricity demand. Electrical frequency is the measure of balance between supply and demand. When supply and demand are balanced, the electrical frequency is at 50 Hz. We must maintain this balance on the system all day, every day. The normal operational frequency range is 49.8 Hz to 50.2 Hz [Eirgrid].

You can see from the above diagram that the frequency has dropped below 49.8Hz on a few occasions over this period.  As a result, the power station tripped and went offline. This then results in the frequency falling even further. At this stage, cutting demand is one of the few options open to the grid operator. This may explain some of the blackouts on these days.

Interconnector Fault Causes Problems for Wind Farms During Beast from the East

As the "Beast from the East" hit Ireland on the 28th February, things were looking good for wind farmers. The east winds were predictable and constant, unlike the variable westerlies that hit Ireland most of the time. Wind energy became baseload power for the first time on the Irish grid. On the 1st March, the capacity factor for wind was 80%, a power output normally reserved for coal or gas generation. However, a problem occurred on the morning of the 28th. The interconnector to the UK (East West interconnector) tripped out. This meant that surplus wind generation could no longer be exported to the UK. High amounts of wind generation would have to switched off or "curtailed".

Wind generation and forecasted wind during Beast from the East. Note how accurate the forecast was
with one notable exception (see later)

East West interconnector fault on 28th February

A further problem happened on the 2nd March as the storm reached it's peak. Power cuts became a frequent event. Power cuts are inevitable of course during storms and periods of extremely high winds, which is very unfortunate for wind farmers as demand for their product, electricity, is reduced just when their supply is at it's highest. In fact, over the four or five days of the "Beast", demand was relatively normal. This is in stark contrast to the Big Freeze event of 2010 where demand reached over 5,000MW (and wind generation was abnormally low). During the Beast, demand reached a high of about 4,600MW on the 28th February. The periods of highest winds (1st - 3rd march) saw demand reach only 4,200MW.

Demand all time peak 2010 Vs Demand during Beast from the East 2018

Power cuts on the 2nd March

 On 1st March, wind energy was generating about 59% of the total electricity production, one of the highest penetrations ever. However, by the next day, as power cuts became widespread, wind energy was been curtailed by as much as 45%. Nearly 1,200MW of wind was been shut down at 4am. 

The period from 1st to 2nd March was when the storm was at it's most intense in Ireland. 
Wind curtailment can therefore be calculated as the difference between forecast wind and actual wind. 
Forecast wind generation was actually equal to demand at times.

The frequency of electricity, normally static at 50Hz, became erratic during the storm as the grid
operator struggled to manage high wind penetrations. This is from the 2nd March.

Had the interconnector been in operation, 500MW of this surplus wind could have been exported.  Demand, in fact, dropped by 10% on the 2nd March compared with the day before, presumably due to the power cuts. 

These are problems that will only intensify as more wind capacity is added and more and more generators are looking to get a piece of the demand "pie". Interconnectors, like storage, seem like an easy solution in theory, but in practice things are often different. 

The Impact of High Levels of Wind Energy on Conventional Plant

On Wednesday 25th January 2017 wind energy reached a new record of 2,400MW for the Republic of Ireland. This post will look at the impacts on some of the other generation sources.

CCGT ( Gas)

Only three out of eight CCGT ran during the day - two in Dublin (Poolbeg and Dublin Bay) and one in Cork (Whitegate). Presumably, the other 5 plants were paid capacity or constraint payments to shut down for the day.

Wind generation was unusually stable during the day and so Dublin Bay and Whitegate mostly followed demand while the output from Poolbeg had a flat profile. Dublin Bay ran the most efficiently. Whitegate's output hovered between 45% and 60% of it's maximum (or rated capacity otherwise known as load). Poolbeg, on the other hand, operated at about a quarter of it's rated capacity. Operating a CCGT at this level leads to higher specific emissions and fuel consumption, something like driving your car in second gear all the time. 

It would have been therefore preferable to have operated Whitegate on higher loads and take Poolbeg off the grid altogether. The requirement for two power stations to be on load at all times in the Dublin area probably lead to this less than ideal situation.

Coal 

Two out of the the three generators at Moneypoint operated for the day and like the CCGT mostly followed demand.   They operated at an average of 50% of maximum output with minimum output at 40%. A load of 40% capacity is likewise not exactly ideal in terms of efficiency. 


Peat

The three peat power stations are being run on baseload and as a result are not affected by high wind levels. The biomass component seemed to be out of action. The question arises as to why (like wind) peat still receives a subsidy if it is always allowed to operate in the electricity generation market (Though I think the peat subsidy is being phased out). 


OCGT (Gas)

Both open gas cycle units at Sealrock operated at close to full output for the day as like wind they have priority dispatch in the system.


Demand Side Units (DSU)

One relatively new problem for Eirgrid is that despite having all this additional generating capacity in place, matching supply with demand is not as straightforward as previously thanks to the presence of stochastic wind energy. Fast acting plant is one answer to this. Another solution is reducing demand during periods of high demand. Demand side units reduce the demand during peak times giving industrial users a choice to shut down production or use their own diesel generators. 

As more wind is added, more reliance will be placed on DSUs and ironically diesel generation. During this day, on average 18MW per hour of DSU was called on to help keep the lights on. Not a significant amount at this stage. But according to Eirgrid :

The capacity of Demand Side Units in Ireland has increased to 230 MW, and is set to increase further. 

East West Interconnector to UK (EWIC)

The UK interconnector played a crucial role on this day. For most of the time, Ireland sent across it's surplus wind but for an hour, between 6pm and 7pm, Irish generation was insufficient to meet the rise in demand as people arrived home from work and turned their kettles and cookers on. It's an unfortunate fact that you can't "switch the wind on". Also you can't simply switch a large power station like a CCGT on. 

The preferred solution by Eirgrid, presumably because it was cheaper than the other option discussed below, was to reverse the direction of electricity in the interconnector. However, the UK was strapped for generation at this time and (incredibly) France were reliant on UK imports. A precarious situation for the UK but 120MW of spare power for Ireland is not a significant amount for a system of their size.

There is still a shortfall of about 200MW at peak time (where blue line is higher than orange line in the second graph above) which I'm not sure how they made up. Possibly more hydro or DSU or some other peaking plant that I may have missed.

The other option available to Eirgrid would have been to simply increase output from Poolbeg as the demand began to rise. After all, it was being operated at well below optimum efficiency as discussed above.  This is presumably what would have happened if there was no interconnector. Usually the low cost of power purchased from UK would make importing a cheaper option but I can't imagine a grid with such a tight capacity margin as the UK's giving away low cost power during peak demand times. But I can only assume it was somehow cheaper in this case.


The power UK sent to Ireland was made up mainly of gas and coal generation with some nuclear and wind : 

The generation mix for the day is given below  :