Saturday, 23 May 2015

Emissions rise during March with increased use of oil generation as back up

Admin Note - There are a lot of diagrams in this blogpost once again but I believe they tell an interesting story so please bear with me. Apologies if you experience formatting problems - these are not intentional !

However, it is an unfortunate fact that the contribution to adequacy of additional amounts of wind decreases progressively and tends towards zero [ESB 2004]. 

Diagram 1 - All Ireland average wind penetration levels (Eirgrid)

We can see from the above diagram that average wind penetration for the month of March has nearly doubled since 2012. Let's see what impact this has had on the running of our electricity system. 

Diagram 2: Fuel Mix March 2012 (wind penetration 13%)

Diagram 2 shows the fuel mix for March 2012. Black represents coal, gas is yellow and green is wind. 
Gas is acting as back up to the intermittent wind. Given Ireland's generation capacity, this is the most 
efficient and cleanest form of back up. Hardly any oil generation was used.

Diagram 3 - Fuel Mix March 2013 (wind penetration 17%)

Diagram 3 shows a similar fuel mix as 2012 but with some oil generation (red shading at top)

Diagram 4: Fuel Mix 2014 (wind penetration 21%)

Diagram 4 again shows a similar fuel mix for 2014 but with small amounts of distillate 
(i.e. diesel) oil generation (light green shading at top) and heavy fuel oil (red).

Now we come to March 2015:

Diagram 5: Fuel Mix 2015 (wind penetration 24%)

You probably have noted that gas generation has become comparatively less and less 
each year as wind penetration increases. But what we see now in 2015, with average and 
maximum wind penetrations of 24% and 61% respectively, is significantly more distillate and 
conventional oil generation. This meant that emissions from conventional generators 
increased as "dirtier" inefficient oil replaced "cleaner" more efficient gas generation. 
Why did this happen ? Well, if we take a look at forecast and actual wind generation 
for a period in March 2015 it will give us a clue :

Diagram 6: Wind forecast and generation March 2015

The intermittent nature of wind is evident in Diagram 6. The red line shows forecast wind and 
it is clear that actual wind (blue line) failed to meet forecast wind on numerous occasions 
during this period.

Oil generators have a unique characteristic in that they are very fast acting, in Ireland it 
typically take eight minutes for them to reach full capacity, compared to say a gas generator 
which can take up to eight hours to start. But there is a trade off - oil produces more emissions 
due to its energy dense nature while gas, once the generator is up and running, produces about 
30% less nitrogen oxide and carbon dioxide than oil. Gas plants are also much more efficient 
in terms of fuel consumption. So what has happened is that fast acting oil generators are 
stepping in to meet loss of supply due to unforeseen drops in wind power.

If we take a system with lower levels of wind penetration, like in 2012 / 13, we can see that 
gas generators can cope with these wind levels as sudden loss in supply from wind generators does 
not cause a major problem to the system. But we can see in Diagram 6 losses in wind generation 
of up to 400MW, which is akin to the loss of the largest power plant in Ireland. One might ask, 
but surely, there is reserve there for such a loss of power - well there is, but in my opinion, 
this would be reserved for the loss of a power plant rather than loss of wind power.

The conclusion from this is that the system can cope with wind penetration of circa 20% but as 
you go above this level, the benefits from wind energy diminish, as you have to back it up with 
fast acting higher emitting plant. I have long believed that we have reached saturation 
point with wind energy and this data confirms this. It is clear that an all wind strategy does 
not make sense.

While nuclear should be an option but requires a long term plan of itself, there is a simpler 
solution, that does not require back up oil plants, new pylon infrastructure or a new expensive 
Grid Code to accommodate high levels of unstable wind energy, to meeting our renewable 
targets - biomass. 

The below presentation gives a good summary of the benefits of this option :

While wind provides non dispatchable generation (incapable of been switched on when 
required), biomass provides dispatchable generation (can be switched on as required). 
This means that biomass generation can replace an existing power station (eg Moneypoint 
coal power station) and utilize existing grid structure. 

Other Data

For completeness sake, the below diagram shows March demand for the years 2012  - 2015. 
You can see that there was a couple of days where peak demand was higher in 2015 
(and also lower) but in general, demand was roughly the same and therefore does not 
account for the increased use of oil generation. 


  1. Another excellent and informative piece by the Irish Energy Blog. The issue with limited modulation of CCGT plant brings to mind the analogy with a milk pasteuriser. Milk is stored at temperatures below 4°C to minimise bacterial growth, and has to be pasteurised by heating it to over 72°C for 15 seconds.
    A simple pasteuriser consists of two heat exchangers; one using steam to heat the milk to 72° and another using a chiller to cool it back to below 4°. A lot of energy is used to heat the milk, and an equal amount of energy to cool it down again.
    This kind of pasteuriser is very easy to operate. If there is no milk coming in, simply flush it with water, and switch it off until it is needed again. Unfortunately, the flexibility comes at the price of low efficiency.
    A modern regenerative pasteuriser features a third heat exchanger in which the cool incoming milk is heated by the hot outgoing milk from the steam heat exchanger, and this has the effect of cooling the outgoing milk towards the 4° target. The efficiency comes at the price of low flexibility; their throughput can only be changed over a limited range, and slowly at that to avoid instability.
    Regeneration ratios of up to 95% are achievable, which means they use just 5% of the energy of a non-regenerative unit. It is rather like driving a car with a block on turning the steering when more than 30° in either direction.
    Ireland has a more than adequate stock of high-efficiency, low-emission CCGT generation plant available, but the efficiency can only be realised under relatively predictable load conditions. The required plant is dispatched with sufficient lead time to meet the predicted load.
    Enter wind generation, which has priority dispatch and is often outside the predicted capacity as illustrated so graphically in the blog.
    Adding erratic generation capacity to a grid is exactly the same as erratically shedding load. The high-efficiency CCGT plant simply cannot react fast enough. The more wind farms are connected to the grid, the higher the proportion of erratic generation, and very soon the CCGT plant has to be shut down and replaced by an inefficient, high emission open cycle generating unit.
    It is the equivalent of winding Irelands milk processing plants back to the kit used in the 50’s because it is no longer possible to plan a continuous flow of raw milk.
    And we call this progress?

  2. Sean O'Dubhlaoigh25 May 2015 at 11:45

    The ASSUMPTION of all involved the wind energy business? that all wind turbines last 30 years and have no output decrease over their allegedly productive lives. Is invalid. This can only be observed in low capacity wind turbines. The closer the operating capacity is to 20 kilowatts the better. If you want long lived wind turbines with no or very little output drops. In a study on the performance of large onshore UK wind farms Dr. Gordon Hughes found that there was a 50% capacity factor drop every 6 years. At year 12 the expected capacity factor of a wind turbine will be 25% of its original capacity factor. From the late 90's the operating lives of wind turbines decommissioned in Denmark dropped dramatically . From 2000 to 2014 , in Denmark,the average operating live of wind turbines commissioned and decommissioned was 6.2 years,. Accompanied with equally dramatic output falls. It will be virtually impossible to factor these realities into forecasting software as these performance drops are closely related to design failure issues like premature component failure and variable operating environment issues. Which are going to be specific to each wind turbine chosen for each wind farm , the environment in which they operate and wind farm layout issues. Until these issues are addressed Eirgrids forecasting is going to exhibit significant forecasting failures of wind energy output. But then again admitting these issues to be major problems would undermine the whole wind program.

    1. The sad thing is that your pension fund and mine have invested up to their necks in these things. So not only do you get a messed up landscape and sleepless nights, but you also get a reduced pension as the losses accrue. But cheer up, the developer who put through the planning and sold on the operation now has a difficult decision to make on the 5th green in Malaga.

  3. Somewhat tangential but still related, the Irish Times reports that Apple are looking for renewable energy suppliers for their Athenry mega data centre.

    In particular
    "The technology giant said proposals must be eligible for subsidies under the State’s REFIT program, which applies to onshore wind, hydro and biomass. Apple only wants projects within the Republic of Ireland, and they must be ready in 2017 or 2018. Those in “proximity to Athenry” will have an advantage."

    Does this mean that domestic electricity users in Ireland already paying some of the highest prices in Europe per KWH will end up subsidizing the power needs of the richest company in the world?

  4. The Chinese conventional suppliers are baulking at wind energy.

    Turbines can be divided into direct drive and gear drive. I fail to see how a direct drive alternator can perform @ 20 rpm. Preliminary results show a capacity factor of 15.45% for Mount Lucas direct drive wind farm. A decent alternator needs a speed of at least 1,000 rpm. a step up of 50 times. A step up gear train absorbs most of the in-putted energy. Heavy steel cog wheels and bearings (must be lubricated), take a lot of blade power just to rotate and I believe transmission resistance rises with torque increases. I estimate they would be lucky to get a quarter of blade power to the alternator. Reading through some publications, it seems they tried to get over the absorption of torque caused by steel bearings, shafts and cog wheels by substituting lighter materials for steel. Titanium is the only one I can think of and it is incapable of being hardened through heat treatment. Parts are breaking. If I were asked to make a step up gear train above 5/1 using steel, I would refuse due to forecasted poor transmission. If i were asked to make it out of titanium, I would refuse due to low impact resistance. The only way would be to use steel inserts on the shaft and steel ring gears on an aluminium or titanium centre. This means joints! Like the ring gear of a car flywheel. I believe these problems get worse as the machine get larger and output increases. My tractor has a 90 hp engine giving 67 kw. A 2.5 mw alternator demands 37 times the torque from the gears and blade as my tractor can provide at best revvs. Can't be done I would have said, then again the object is to make money, not electricity.