CO2 Emissions Variations in CCGTs Used to Balance Wind in Ireland

There's an interesting analysis of CCGT emissions and how it is affected by high wind generation along with an interesting discussion posted over at Energy Matters website :

http://euanmearns.com/co2-emissions-variations-in-ccgts-used-to-balance-wind-in-ireland/ 

The Electrification Paradox

It is hard to determine exactly what the Government's long term strategy for electrification is. Like a lot of their energy policies, it is very incoherent and full of inconsistencies. Take Eirgrid's new initiative to reduce peak demand by compensating households to switch off :

EirGrid has launched a competition to identify a company who will pay householders to reduce their energy consumption at high demand times. This is one of the first times a transmission system operator (TSO) has run such a competition.  This pilot project will benefit up to 1500 householders, who will see their annual electricity bills cut by up to €100 by participating in this scheme.The competition is open to companies who will work with homeowners and EirGrid to provide a service known as “demand response”. Demand response enables electricity customers temporarily reduce their electricity consumption in response to requests from EirGrid, resulting in savings on their bills.In Ireland, businesses and industry can already do this; however, it has not been available to homeowners until now.The goal for EirGrid is to manage demand on the national grid and give homeowners more control over their electricity bills. Fintan Slye, Chief Executive, EirGrid, said: “This is a really exciting development in the electricity market. We often hear about “Smart Grids” and their benefits – this is it in reality - technology innovation that puts money back in people’s pockets.  Any measure that helps reduce overall demand on the grid, while not affecting our daily lives at work and at home, can only be a good thing in the long term.”

The reason they need to do this is because so much of our electricity is set to come from intermittent sources, mainly wind. If you have an intermittent source, then demand must also become intermittent i.e. dependent on the weather.

But the Government's Plan is to increase rail electrification in the future :

Rail electrification substantially reduces the use of fossil fuels in public transport. There has been significant progress with the introduction of DART and LUAS, and the recently published Capital Plan 2016-2021 [40] provided for further such public investment in the Greater Dublin Area. Further rail electrification will be a priority in future capital plans.

This will have the opposite effect, substantially increasing demand for electricity during peak demand times when people are leaving work. So people travelling home on an electric train will get paid not to switch on their cookers and other appliances when they arrive home. But the demand due to increased rail electrification will far outweigh the reductions from Eirgrid's demand side measures. This means that more dispatchable power stations will be required, most likely from fast acting fossil fuel sources such as gas or oil, to ensure that when a train is due to depart, it actually does so.

As noted on this blog before, no proper analysis has been done on this in terms of emissions or fuel saved. If we ran more efficient gas plants, with gradual ramping to meet gradual increases and decreases in demand, would we have more or less savings than one with a system that used wind energy and fast acting inefficient gas and oil plants as back up ?

And if you took the billions been spent on new energy infrastructure and invested that in passive houses and energy efficiency, you would most likely have a lot more savings than the current plans.

But if you don't do the calculations and allow ideology take over, as during the Celtic Tiger, then you are doomed to fail.

Storm Desmond - the problem with too much wind energy in one graph

Forecasted Wind energy output for 6th December, the day after Storm Desmond, shows a loss in wind generation of 91%

As I write, Storm Desmond is raging outside my window and wind energy output is providing almost 50% of demand. It is coming at a price, however, with ESB Networks reporting that about 12,000 customers had to go without power today due to damaged cables. But electricity is not just required on a Saturday, it is required every day. According to Eirgrid forecasters, wind energy output is forecast to drop by as much as 91% tomorrow, from nearly 2,000MW to 160MW over 16 hours. (a drop in capacity factor from 87% to just 7%)

This explains why wind has such a low capacity credit. Let's look at what Eirgrid had to say in 2009 :

However, the benefits [of wind energy] tends towards saturation as wind penetration levels increase. This is because there is a significant risk of there being very low or very high wind speeds simultaneously across the country. This will result in all wind farms producing practically no output for a number of hours (note that turbines switch off during very high winds for safety reasons). In contrast, the forced outage probabilities for all thermal and hydro units are assumed to be independent of each other. Therefore, the probability of these units failing simultaneously is negligible [Eirgrid 2009].

What they mean is that a gas power plant (CCGT) of 400MW might drop out but will do so independently of another gas plant. So that would be a loss of 400MW which standby reserve would adequately replace. But wind farms do not act independently of each other.  Instead, when one drops out, chances are the rest of them will aswell. So now you have a huge hole in generating output of about 1,900 MW, equivalent to about 50% of tomorrow's demand. So you need to have about 5 large gas plants ready to step in. Starting these type of plant up from "cold" is not a good solution as engineers have stated that this is 20 times more damaging to plant than "warm" starts. It is also very expensive and high emitting in pollutants. So gas plant are kept running on low load behind the wind and will then step in tomorrow to pick up the load as wind slackens off. 

Eirgrid are lucky in that the wind is forecast to decline over a period of 16 hours. What would happen if they are wrong and it falls off at a quicker rate ? Then they may need more fast acting plant like diesel or open gas cycle turbines. These are more polluting than efficient CCGT type plant and will result in higher emissions, thereby negating some of the benefits of having all this wind generation. 

Can we close down any of our power stations and replace it with a huge fleet of wind turbines ? Can we really make a transition to a more sustainable "green" based energy supply ? 

As you can see from the above, the answer is no, not with wind farms.

Five Reasons why we have reached saturation point with wind energy

To any impartial analyst, Ireland has reached saturation point with wind energy and it should now be time to put a pause on new wind development and consider our options. No damage was ever done down through history by pausing before deciding what to do next. Think of how many billions of euros we could have saved if this was done in 2006.

1. Dumping of wind power and the 50% limit on wind - recent evidence shows that during periods of high winds we have to dump more and more of available wind energy to maintain a safe secure supply of electricity.  On the 30th March, at least 26% of available wind energy was dumped. 
2. Over capacity - We now have generation capacity equivalent to double our peak demand and three times that of our average electricity demand needs. Let's use up this excess capacity before we start building any more. No new generating units (including wind) need to be built unless they are replacing retired units.
3. Baseload plant minimum load requirements - there is a requirement for 5 large generating units to be running at all times for "dynamic stability". These comprise combined cycle gas turbine plants and Moneypoint coal plant. This means they can never be completely switched off. Increasing wind penetration further will exacerbate the inefficiencies inherent in running these plant on low loads, thereby negating any additional savings due to adding more wind.
4. Electricity bills are one of the highest in Europe - government policy has locked society into high electricity prices with the preference towards subsidized forms of generation meaning savings from falls in wholesale prices can never filter down to consumer's bills. Another factor is that an over supply of generation capacity results in units requiring subsidies and capacity payments to recover their high fixed costs as payments for energy generation become insufficient and staggered due to low demand and more intermittent wind on the system. There are also extra costs due to new infrastructure required to carry the wind power.
5. Impacts on other sectors - The tourism and equine industries are two of the largest industries in Ireland supporting many direct and indirect jobs. Chances are if you live outside any of the main cities, your job is dependent in someway on either of these industries.  Planting wind farms and associated pylons near scenic and horse breeding locations will have a negative impact on these important industries.  The Irish Hotels Federation recently warned that the location of energy infrastructure should not diminish the natural beauty of the landscape because this is an important element of the Irish tourism product. Already, this impact is being felt with one castle owner recently saying "The tourists can't believe it. They said we're mad. They said we're ruining our heritage. They say it's disgusting to go around Ireland now"

Dublin Electricity Generation - An Analysis Part 2

In Part Two, I will take a closer look at Poolbeg and Dublin Bay CCGT over the same period as Part 1 - 1st November to 4th November, 2014. In Figure 1, these are the green and blue lines along the
bottom :

Figure 1: Total Wind Generation and Forecast for the Republic (Eirgrid) and Profiles of Dublin Generators (SEMO) 
1st Nov - 4th Nov, 2014

Before I proceed, a little explanation on the operation of CCGT (combined cycle gas turbine) is required. This type of plant is the most efficient for converting gas into electricity. It is basically a gas turbine, such as what is in an airplane. But unlike the gas generators of old where the steam is released into the sky through a stack, the CCGT converts the steam into electricity aswell. We can see from the following graph what the efficiency is like at various output / loads:

From http://www.thermalplant.com

So for instance, a plant on full load producing 400 MW, a typical Irish CCGT, will have an efficiency of about 58%, but when throttled back to 200 MW, the efficiency has dropped to 50%. While the curve is not continued below 50% loading on the plant, it is in fact a steeply dropping curve, such that at 100 MW (33% load), the efficiency is less than 40%.

In other words when running a CCGT at 33% load you need to burn 2.5 MW of gas to generate 1 MW of electricity. If you ramp back up the power station to 100% load, you only need to burn 1.7 MW of gas to produce the same 1 MW of electricity. This means effectively that by operating the CCGT at one third of its design load, the unit gas consumption has gone up by 50%.

The same can be seen in the carbon dioxide emissions in Figure 2. When the CCGT power plant is operating at full throttle, the carbon dioxide emissions are as low as 0.35 tonnes per MWh (350 g/kWh), but start to rise rapidly as the output of the plant is reduced. 

From http://www.thermalplant.com

So bearing this in mind, we now come to Dublin Bay and Poolbeg.

Dublin Bay CCGT

Dublin Bay is a 415MW gas plant. Figure 2 compares the load profiles for Dublin Bay for a period with no wind in October (9th to the 12th), and the above high wind period in November.

Figure 2: Dublin Bay operating at different loads

We can see that it was operating at close to full load, producing around 390MW (95% load), during the period in October. According to a recent EPA report, Dublin Bay has an efficiency of 56.97%. So to produce about 390MW of electricity, 680MW of gas was consumed. In November, when high winds moved across the country, we can see that there was alot more cycling, particularly in the first two days. When it ran at 50% load in November, its efficiency was about 50% or slightly below. So to produce 200MW, roughly 400MW of gas was consumed. So there was a fossil fuel saving in the region of 280MW when operating at 50% load as opposed to 95% load, but there was also a corresponding rise in emissions of circa 0.05 tonnes per MWh. These savings don't take into account any increase in reserve that may have occurred during the period of high winds. This information is unavailable unfortunately.

Engineers in the trade are now realizing that this increased cycling is not something that is without consequence. Gas power plants evolved over the years from pretty lousy efficiency to a very high degree of efficiency in modern times due to the great ingenuity of engineers. These advances are now been undone by the high levels of wind energy been allowed into grids across Europe and elsewhere. As posted in a previous blog , German engineers are now claiming:

"that existing plants are not technically laid out for the operational requirements of today, which naturally is being altered due to the highly intermittent input of increasing amounts of solar and wind energy on to the grid. This rapid increase in renewables in recent years in Germany has put operational demands on existing gas and coal power plants, which are simply not technically designed for it. The plants must be more frequently switched on and off in order to be able to compensate for the fluctuations, which are associated with electrical inputs from sun, wind and water. The degree of load change is partly more than 200 times higher than that permissible for the power station. As a result the danger of lasting damage to the power  plants grows – along with increasing risks to the security of electrical supply."

So who will pick up the tab for the increased maintenance of these generators that will result ? Well, ESB, a semi state company, owns Dublin Bay, so the answer is the Irish taxpayer.  The Irish people part own a very expensive asset in Dublin Bay, one of the most efficient plants in the country at generating electricity, now been run increasingly like its inefficient ancestors of old. It's akin to buying a brand new car and driving it in second gear on a motorway. So there is a trade off between the fossil fuel savings from wind and the inefficient operating of plant. Of course, if wind could replace the plant in its entirety, then the problem would be solved.

Poolbeg CCGT

Poolbeg is a 463MW gas plant also operated by ESB. However, it is not a modern CCGT like Dublin Bay, but rather an older model. Therefore it has a lower efficiency - 46% according to EPA. Figure 3 shows the different load profiles during the period of high winds in November and the same period above in October with low winds :

Figure 3: Poolbeg CCGT different load profiles

During the period in November, Poolbeg was ramped down to 25% to allow the high wind penetration into the system. Its normal position is seen in the blue line during the period in October, where it sits at circa 50% and ramps up when there is additional demand. Because demand has fallen in recent years, there is no longer a need for all the extra capacity in the Dublin region and so Poolbeg (which has to run to maintain voltage) runs on half load most of the time. (by the way, average Demand over the two periods in this study is almost identical). So during the low wind period, the efficiency of the plant is circa 42% based on an average load factor of 58%. This results in average fuel consumption of 550MW to generate 230MW of electricity. But during the period of high wind, with a load factor of 25%, the efficiency of the plant has now fallen off a cliff, exacerbated by the fact that Poolbeg is not an efficient modern CCGT plant to begin with. So the efficiency in this case is circa 20%, meaning that fuel consumption is in the order of 574MW to generate 115MW of electricity. So the plant actually consumed more fuel when the high winds moved over the country - See Figure 4.

Figure 4 : Poolbeg CCGT fuel consumption - ramping down to low loads can lead to increased fuel consumption and
CO2 emissions

Likewise, the CO2 emissions were greater during the high wind period which brings up the issue as to  whether the EPA should be monitoring and doing something about situations like the above. 

So it can be shown that very high winds can lead to an increase in emissions and fuel consumption -   the very opposite of what was intended - what is known as "the law of unintended consequences". 

While there was a fuel saving in Dublin Bay (when one discounts the emissions involved in the         installation and manufacture of wind turbines), there was fuel and emissions cost in Poolbeg.

There is also potential maintenance problems that may result for both plants - with ESB and the           taxpayer / consumer picking up the tab.

This is another problem that will be exacerbated when more wind is added to the system. As already   pointed out, two of the large gas power plants in Dublin have to run at all times and can't be shutdown regardless of how much wind energy is generated. So these plants will be forced to increasingly ramp down to even further lower loads, further increasing fuel consumption and emissions. This proves that wind energy is only of any use at low penetration levels unless there is large levels of hydro in the system such as in Scandanavia (and even there, they have to export a large proportion of their wind output).

Aghada (AD2) Combined Cycle Power Station

Aghada (AD2) in County Cork is a 431MW Combined Cycle power station operated by ESB. The diagram below (Fig 1) shows how central it now is to power generation in the South West region since Ireland’s rush for wind energy. In 2009, 920MW of wind was installed. By the end of 2013, installed wind capacity was at around 1,772MW. This has resulted in more frequent ramping of this generator.

The second half of the month is the most suitable for comparative purposes. Back in 2009, the plant was ramped up to meet peak demand and then ramped down to exactly 50% load for extended periods - the threshold below which significant inefficiencies begin. It is well known in the industry that running a gas generator at below 50% load results in more fuel consumption and consequently, more CO2 emissions being emitted than if the plant were running at full load. But in 2013, the plant was used to accommodate even the smallest variances in wind generation, hence the sawtooth edge. De-ramping was to between 48% and 50%, teetering over the inefficiencies cliff. The rugged irregular edge of 2013 compared with the smooth fine edges of 2009 proves that wind is many more times variable than demand, contrary to what SEAI claimed in their recent report “Quantifying Ireland’s Fuel and CO2 emissions savings” (Wind generation variability in 2012 was less than electricity demand variability - Page 9 of SEAI Report )

Interestingly, this inefficiency threshold of 50% is the same as the threshold for non-synchronous wind penetration and interconnector imports, meaning that wind generation and imports can be accommodated in the system up until the point where they reach 50% of demand, without causing large scale thermal inefficiencies. So the question is, what will happen when Eirgrid try to integrate 75% wind penetration into the system as is being investigated under their DS3 Programme - a requirement that will be necessary to meet renewable targets. This could result in de-ramping of gas plants below even 30% load where fuel consumption and CO2 emissions are circa three times what they are at full load, thereby completely negating any fuel / emission savings due to wind energy. This situation would also be bad financially for gas plant operators as their plants would be consuming more fuel but receiving less from the market. Its doubtful that they would be allowed to close down though, which means more subsidies in the form of capacity and/or constraint payments.

The graphs below show Aghada (AD2) and two large Cork wind farms - Coomacheo and Boggeragh. Together these wind farms total 116MW. Fig 2 shows actual output while Fig 3 shows Aghada’s output adjusted to make the comparison easier. It can be seen that in the early and latter parts of the month gas generation was used to fill gaps in supply due to calm winds. According to Eirgrid, Aghada (AD2) had a 93% availabilty in December, so its shut down “cold” status during the middle parts of the month indicate displacement by wind and possibly coal. Fig 4 shows Moneypoint (MP3) and Aghada (AD2) lined up. The erratic profile of both generators is notable in the second half of the month, due to increased wind penetration in the system. Indeed, on 17th December, a new wind output record was set of 1,769 MW. Aghada’s “cold” status may also have been due to operational constraints in the Cork region, where the maximum thermal generation at any one time is restricted to 1,100 MW. (Eirgrid Operational Constraints, March 2014)