Blackouts loom this winter

EP UK Investments, 80% owners of the Tynagh power plant and two plants in Antrim, took a high court challenge against the contract.

With it not due for an initial hearing until October Eirgrid appear to have scrapped it.

https://www.irishtimes.com/business/energy-and-resources/uk-energy-firm-challenges-eirgrid-s-awarding-of-contract-to-rival-1.4639658

Renewable power is what has us in this mess. Look at how many System Alerts issued since 2018 and then cross check what the wind and solar were doing on those days. Zilch is the answer so an aging thermal fleet is required to pick up the slack.

They are clueless! In addition to this they (the regulators mostly) should be rewarding consumers who are willing to shift their load from day to night or from non-windy to windy times. I charge my car with my solar PV during the day when the sun shines - even though I should be incentivized to charge at night the national load is lower. Simply because I get nothing back for exporting to the grid. So might as well use it than giving it away for free. But net net this is increasing daytime loads and decreasing nighttime loads while we should be doing the opposite.

Also never understood why the standing rate for a D/N meter is higher. Talk about incentives.

That's possibly because you are posing an impossible question. There is no such thing as "excess power" at present. It is generated as it is consumed or it is stored in the Hill. All the batteries on the system are merely providing system services and aren't there to give medium term capacity or balancing benefits.

By "excess power", I'm assuming you mean that which is curtailed or constrained if locational specific? However, for any curtailed asynchronous energy to be utilised, you also need 30% (or 25% during the upcoming 75% snsp trial) coming from synchronised sources. That means additional conventional plant committed and dispatched upwards.

It is correct. SNSP is 70% so wind can be a little higher than that, depending on what other generation is on and exports from the system.

Excess wind isn't some magical energy that can be put in a bottle and delivered when and where you want it. It still has to use the same transmission and distribution system as all other electricity and hence is bound by the snsp limit, like all other generation on the island. There are a minimum number of conventional generators online at all times to ensure system stability. If wind is high, these may be at their minimum output as long as the SNSP level doesn't exceed 70%. If it does, the wind/solar is curtailed and turned down and the conventional plant is moved up to maintain the frequency. Interconnector flows are slightly different due to the market closure times but they also influence the SNSP.

Also, in high wind, conventional generators are often brought on to minimise constraints by pushing the power on the right side of the problem. Look at what SONI do in relation to Moyle when there's wind in the SW of Ireland, they push up on Ballylumford, Kilroot etc. It's so hard to model but their explanation previously was that this is the consequence of not building the second North South line. It would be great to see a published network model on this so we could really understand the constraints issue.

I suppose the question then is at what % SNSP does wind start getting curtailed. If it's at 65% then we have room - assuming no constraints - then there is "excess power" that can be used. If you look at this report http://www.eirgridgroup.com/site-files/library/EirGrid/Annual-Renewable-Constraint-and-Curtailment-Report-2018-V1.0.pdf SNSP was not the reason for curtailment majority of the time. All of those times there was excess power that could have been used.

The other source of slack are the interconnectors. We could export less and use more internally if we had dispatchable loads.

I'm betting the data centres, such huge consumers of energy, won't be getting hit with any blackouts.

This comment really epitomises how little most people understand about the power system and the market. Yes, sometimes wind is curtailed for system reasons before the SNSP limit is reached, for example, overnight when all conventional plant (less any reserve constraints) have been turned down to their minimum and the frequency is increasing. So what? You increase demand a little till the SNSP limit is reached, thereby not having to match with increased conventional output? Great in theory but if there were signals to do so, you'd need some sort of feedback because otherwise you have a peak at 3am rather than 6pm. Plus consumer behaviour doesn't necessarily want their oven on at 3am or their dishwasher running 6 times overnight (with no refills). What you are talking about is not just a real time price signal, but a forecast real time price signal that is static. However that's impossible since the price is set depending on the generation online to meet the load at a point in time and both are variable.

Demand response is already a thing and DSUs can (and often do) elect to match the system demand through their pricing.

As regards Interconnection, we are part of a wider European market. You don't just get to change flows on a whim. They are determined by prices across the region and flows go from the cheapest to the most expensive areas. Dispatchable loads may not change this because if the SEM is particularly windy (0 bid price) our price would go to 0 or even negative. This drives the interconnector flows, which are set BEFORE any balancing market activity. You'd need to change the entire set of market rules across Europe and for what? So that you can turn up Intel's demand when they aren't expecting it? Or turn on residential storage heating in the middle of the summer? There are only so many levers that can be pulled.

Grid signal? What on earth are you talking about? I work in power system standards and certification so I reckon I have a better understanding of the present and future state of the art than you do, having spent years researching and lecturing on the topic. There is no mythical grid signal, save for the 50Hz supply you receive. PLC technology is largely phased out at this stage and replaced by fiber so unless you are proposing an installation of costly RTUs at every home, business, farm and factory, your "idea" is unworkable.

If you are talking about is modulating some sort of carrier onto the 50Hz supply, you'll quickly run into issues with sensitive loads, harmonics and cable amplification that'll put paid to achieving anything coherent. You'd be better off asking Siri or Alexa to do something.

As far as EVs are concerned, there are far too few of them to impact the potential blackouts this winter, which is what this thread is about. Shifting a few hundred kW to the valley is welcome but its very small in the big scheme of things. If anything, they'll make the problem worse because people use the inefficient heating function in EVs more during winter, often while plugged in and during the morning load rise before they head off to work, which makes for a very unattractive ramp for the system. No point heating a car at 3am for it to be cold again by 7-8am. The same can be said for the return journey at 5pm. As far as bidirectional transfers go, unless battery technology improves, I cannot see many stupid enough to sign up to V2G and erode your battery capacity providing a service for little returns. Especially if the grid runs down your battery just as you need it for an emergency journey. Why invest in an expensive EV if you no longer have control of its state of charge or battery lifetime?

As for interconnector and network flows - some things do have to be set in stone. The power system works based on predictive behaviour. Obviously unplanned outages can affect that and reserves kick in, till the system is restored within operational limits. Generators are brought online against forecast demand. They take anything from minutes to hours to get up to speed, not milliseconds which would be needed for true demand response. Plus, if there is no forecast of generation and guesstimate of prices, how do you propose to create your mythical signal? How do you achieve it post contingency if something trips? There's no AI out there clever enough to do it yet, nevermind in real time.

As for changing rules, as I said, I work in this space. It takes years (decades) to make changes as it requires multi jurisdiction coordination. Everything is agreed at EC / Entso-E level, then filters downwards. Then market systems have to be designed, developed and tested before being deployed. There's lots of stakeholders with lots of vested interests from developers, suppliers, traders, all seeking to make returns and consumers / regulators trying to get value. We've had 2 SEMs in Ireland and I suspect the next one won't be till 2027/8. Again, not much use for this winter.

I don't work for the regulators, I work in standards. Very different things. I could care less about what technologies deliver as long as the power system operates safely and within acceptable limits. It looks like this winter will be a challenge, though its still within design criteria assuming an 8 hour Loss of Load Expectation, its only natural to experience the odd disturbance to supply. However, once RES-E penetration increases, I expect that'll worsen unless new entrants arrive in the market.

Data centres have diesel generators and UPS batteries as backup. They can manage unless there are fuel shortages too.

EI have just announced another hike

Coal is suddenly back in fashion. The amount of coal we have burned to generate electricity this year has quadrupled

Dublin Waste-to-Energy Facility - Wikipedia

Yeah, it's in the report allright.... (thanks for that garo)

The median demand forecast predicts data centres accounting for 25% of all demand by 2030. This increases to 33% in the high demand forecast. This represents a current forecast position; it does not include all of the most recently received data centre applications seeking connection, discussed below

Pretty much inline with my own estimations. To have datacenters using 70% of the countries generation.....just not possible. Not even remotely.

That said, as much as 25-30% seems a lot......I always get concerned when I see politicians saying things like "Ohh we'll have a moratorium on building new datacentres" etc. rather than fixing the grid to support the infrastructure and getting the large consumers to help pay for it.

Not building datacenters, means no construction jobs in building the centers, jobs to support the infra, revenue (rates/taxes) from the companies. Back in the day the large consumers were like AUGHINISH ALUMINA. To tell them to "look elsewhere" is ..... shortsighted.

Very true, and I speak as a software developer, the number of people to run a full datacenter is probably 100-150. You'd think it'd be more, but it's not and mostly that's in the facilities side of things. Maintenance of AC, (heating/cooling), electrical, etc. only a few of the hard techies youd expect to see. Still if you have companies willling to spend €250m on construction of facilities and you can get them in the door, that's jobs for what 500-1000 people over 3-4 years?

The Aughnish example I gave is a good one. They setup their own 140mw generation plant to help the grid out. This was 20 years ago and back then they would probably have been the largest single consumer on the grid for 20 years since they were setup in the early 80's. They were the "datacenters" of the day.

Aughinish 140mw power plant welcomed by ESB - Independent.ie - article from 2002

As you say, it'd be great to see the tech companies doing the same, starting out at the least covering their roofs with PV, although it would only be "a start" in the right direction. Their power requirements would need sq km's of roof!! But if they could get 5% from their own roofs, hell it's helping at least. Get some windmills cofunded with the government (they maybe doing this already? - I don't know).

Bugs the bejasus out of me when you have politicans directing the policy when they haven't a clue. Not a clue.

Didnt all the dairy plants build gas CHP plants in the mid 2000s as well ?

Look at all the gas line extensions ,from that era ... macroom , ballineen , nenagh ,charleville ,listowel

As well as chps in mallow and mitchelstown ,and I assume every other large milk processing plant ...

These plants are big users of heat - and electricity , so it's more cost effective to just buy the gas themselves,

I'm assuming these Are all the simpler , but less efficient open cycle plants ..

But anyway datacentres don't want heat , in fact they're trying to get rid of low grade heat. .

They would get way more that 5% from a roof plastered in PV.

hmmmm, initially I thought the same until I looked into it and then ran some numbers in my head. Your typical datacenter wolfs down 60-100mw. Fairly stable too, but they do reduce overnight and at the weekend slightly (10-20%) The biggest solar farm in the UK (Shotwick solar park) produces 72mw and that takes up 250 acres. 250 acres!! Yikes!

The UK’s 5 Biggest Solar Farms | Community Solar | Deege Solar

So assuming they had that size it would match the load, but only when the when the sun is shining. In reality the datacenter roofs wouldn't have anywhere near that kind of surface area, and of course we know from our own residential production (sept being a good example) that the sun is sometime a fickle mistress - and in winter, it's a token gesture at best. No, 5-10% at a push I'd say with all the roof plastered in PV. I would very much argue that they should plaster the hell out of the roof, but sadly that's all it will cover of the load. But they could use that to charge batteries and do peak load shaving etc. That's really the main issue for Eirgrid as I understand it.

I like your ideas though. I'd add one more.

1. Get them to co-fund the installation of the wind turbines. If they are using 100mw, get them to pony up say 30% of the cost of 20x9mw turbines, and then sweeten the deal by giving them preferrencial rates for the units. Or even for free for 5 years or something. Essentially paying up front for the leccie. Ireland benefits from having the infrastrucutre and they benefit from cheap(er) leccie.

Dunno if I'd go as far as to say short sighted - I mean they operated for 20 years nearly only shutting down last year when the got refused permission to keep burning peat. 20 years is a good crack of the whip, no?

Personally I'd have preferred to see other alternative generation facilities, but solar was nowhere to be seen back in 2002 and wind was only starting to gain traction. Not sure what else they could have done - but to kind of circle back to the original point of the black outs in Ireland this winter, it's the great problem with renewables isn't it? When the sun don't shine or the wind doesn't blow.....we're forced back into firing up gas/coal/oil stations. :-(

Yeah, but all the viable places for hydro have been used. There is that plan in the silvermines I came across, but I haven't looked it up to see what they were planning.

Don't laugh, you can actually convert the kinetic energy in raindrops into electricity via piezo electric cells. It's feck all though. Like tiny tiny, even in a storm.

(PDF) Piezoelectric model of rainfall energy harvester (researchgate.net)

Once they nail that lads......Ireland with be the new "Saudi Arabia" of the energy world :-)

Isn't that essentially what these "green" DCs do already? They basically pay a wind farm to put an equal amount of energy back onto the grid as they're using already