Who's in the UK and has caved in and bought air conditioning?

There is another interesting angle here: AC can actually be a useful flexible load on a renewable-heavy grid.

We increasingly have periods when solar/wind generation is deliberately thrown away because there isn't enough demand, storage or transmission capacity — curtailment.

AC, heat pumps, water heaters and EVs give us loads that can be shifted into those periods. Pre-cool the house while solar is abundant, then reduce the load later.

Maybe the optimization problem isn't only making AC more efficient. It's making AC smart enough to consume energy when the grid has too much of it.

ggabe said:

is deliberately thrown away

Perhaps 'not being sold' is more accurate here ? It's not as if the power generation companies would be giving it away for free.

Pump storage hydro, synthetic fuel, desalination, pumping water between reservoirs, filtration etc. are all potential flexible loads for renewables that would be beneficial to society.

However perhaps only pump storage hydro is viable here if it is owned by the same parent companies as those producing the renewable energy. But they already know this as they have been doing this for decades with fossil fuel power stations pumping the water back up into the upper reservoirs during off-peak ready for the next peak demand.

ggabe said:

It's making AC smart enough to consume energy when the grid has too much of it.

This has been looked at before under a wider context of 'smart appliances' which was aimed at balancing the load on the grid to reduce the issues of peak demand potentially overloading the existing grid. White goods such as refrigerators and washing machines would automatically make use of the off-peak periods and smart electricity meters would automatically compensate consumers for these devices with reduced energy tariffs for those devices. However I recall that some early attempts showed that the schemes were often short-lived as were the promised reduced tariffs.

I think there is an important distinction between “not being sold” and curtailment.

“Not being sold” makes it sound primarily economic: there is electricity available, but nobody wants it at the offered price. Lower the price enough and perhaps you find a buyer.

The curtailment I’m talking about is different. The generator could produce, but the grid cannot accept the output because of transmission/system constraints or because generation exceeds what the system can absorb. In that case even a price of zero — or sometimes negative prices — doesn't necessarily solve the problem. The potential energy really is being discarded.

And yes, smart appliances and demand response aren't new. What is changing is the optimization target. Historically it was mostly move load away from peaks. With high renewable penetration we increasingly also want move flexible load toward periods and locations where renewable generation would otherwise be curtailed.

That's where AC and heat pumps get interesting. A building is already a substantial thermal store. Pre-cool it slightly when the grid has excess renewable energy, then reduce compressor load later.

So I don't see this as another off-peak tariff scheme. The interesting problem is making millions of existing thermal loads respond intelligently to real-time grid conditions.

michaelkellett said:

I haven't studied the UK smart meter spec in detail but it does have quite extensive messaging capability

Assuming that they can complete the roll-out, get them all working in smart meter mode, fix all the broken ones, and replace earlier generation communication hubs before 2G switch off, then yes 

Oh, and make it robust enough for control messaging purposes, as smart meter comms can still be down for weeks at a time ever since the DCC got involved. I'm not sure how widespread this is however, it may not be significant enough when averaged out.

michaelkellett said:

to have a dialogue like  ggabe suggests via an existing smart meter

Yes, I was thinking more of the interfacing of existing appliances. How does the existing smart meter play nicely with all those legacy appliances that are now too old to be receiving support updates from manufacturers but not old enough for consumers to be considering replacing them with a new compatible appliance.

I guess some of this is starting to be talked about in the UK Clean Power 2030 objective though.

I haven't studied the UK smart meter spec in detail but it does have quite extensive messaging capability. It might well be possible to arrange for a local energy management system to use the smart meter to have a dialogue like  ggabe suggests via an existing smart meter. Then all it takes is one of the bigger players (Octopus spring to mind - they seem a bit more adventurous than some)  to offer the service and it could start to happen.

It sounds a better bet than a lot of the storage schemes that are being pushed by Gov  - most of which are big, expensive, destructive and slow to build.

MK

Yes — a heat pump in London obviously can't fix a transmission bottleneck in Scotland. The flexible load has to be on the useful side of the constraint. That's why I deliberately said periods and locations.

But I think your last sentence is actually the interesting engineering problem. We already have millions of controllable thermal loads, each with some inherent energy storage: the thermal mass of the building, hot-water tank, refrigerated space, etc. What we're missing is a good abstraction between what the grid needs and what those loads can offer without inconveniencing the user.

Instead of the grid saying “turn compressor #12345 off,” an HVAC system could expose something more like: I can absorb an extra 2 kWh during the next three hours and then reduce my demand by 1 kW between 6–8pm, while keeping the building within the user's temperature limits.

Aggregate thousands of those and you have something resembling a distributed virtual battery.

There are pieces of this already — demand response, dynamic tariffs, OpenADR, smart thermostats, EV managed charging, etc. What I find interesting is that we still don't seem to have a widely adopted flexibility interface that makes arbitrary HVAC/refrigeration equipment a grid resource.

That seems like a fun engineering problem rather than just an electricity-pricing problem. :-)

ggabe said:

grid cannot accept the output because of transmission/system constraints

For this particular aspect, then I don't see how AC / heat pumps in the home will solve the problem ?

However it wouldn't prevent local storage and use.

I recall reading about the issues some ultra fast EV charging stations are having getting adequate connections to the grid for their energy demand. They are having to invest in substantial local battery storage solutions to get them through peak demand throughout the day. Off-peak then they are still maxing out their connection to the grid in order to recharge their batteries.

However I suspect that investing in either upgrading transmission or local storage and use is seen as being less profitable in the short term than curtailment (although your Wiki link suggests it has been done elsewhere).

ggabe said:

The interesting problem is making millions of existing thermal loads respond intelligently to real-time grid conditions.

Especially when there are a lack of standards in place to do so.

ggabe said:

grid cannot accept the output because of transmission/system constraints

For this particular aspect, then I don't see how AC / heat pumps in the home will solve the problem ?

However it wouldn't prevent local storage and use.

I recall reading about the issues some ultra fast EV charging stations are having getting adequate connections to the grid for their energy demand. They are having to invest in substantial local battery storage solutions to get them through peak demand throughout the day. Off-peak then they are still maxing out their connection to the grid in order to recharge their batteries.

However I suspect that investing in either upgrading transmission or local storage and use is seen as being less profitable in the short term than curtailment (although your Wiki link suggests it has been done elsewhere).

ggabe said:

The interesting problem is making millions of existing thermal loads respond intelligently to real-time grid conditions.

Especially when there are a lack of standards in place to do so.

michaelkellett said:

I haven't studied the UK smart meter spec in detail but it does have quite extensive messaging capability

Assuming that they can complete the roll-out, get them all working in smart meter mode, fix all the broken ones, and replace earlier generation communication hubs before 2G switch off, then yes 

Oh, and make it robust enough for control messaging purposes, as smart meter comms can still be down for weeks at a time ever since the DCC got involved. I'm not sure how widespread this is however, it may not be significant enough when averaged out.

michaelkellett said:

to have a dialogue like  ggabe suggests via an existing smart meter

Yes, I was thinking more of the interfacing of existing appliances. How does the existing smart meter play nicely with all those legacy appliances that are now too old to be receiving support updates from manufacturers but not old enough for consumers to be considering replacing them with a new compatible appliance.

I guess some of this is starting to be talked about in the UK Clean Power 2030 objective though.

I haven't studied the UK smart meter spec in detail but it does have quite extensive messaging capability. It might well be possible to arrange for a local energy management system to use the smart meter to have a dialogue like  ggabe suggests via an existing smart meter. Then all it takes is one of the bigger players (Octopus spring to mind - they seem a bit more adventurous than some)  to offer the service and it could start to happen.

It sounds a better bet than a lot of the storage schemes that are being pushed by Gov  - most of which are big, expensive, destructive and slow to build.

MK

Yes — a heat pump in London obviously can't fix a transmission bottleneck in Scotland. The flexible load has to be on the useful side of the constraint. That's why I deliberately said periods and locations.

But I think your last sentence is actually the interesting engineering problem. We already have millions of controllable thermal loads, each with some inherent energy storage: the thermal mass of the building, hot-water tank, refrigerated space, etc. What we're missing is a good abstraction between what the grid needs and what those loads can offer without inconveniencing the user.

Instead of the grid saying “turn compressor #12345 off,” an HVAC system could expose something more like: I can absorb an extra 2 kWh during the next three hours and then reduce my demand by 1 kW between 6–8pm, while keeping the building within the user's temperature limits.

Aggregate thousands of those and you have something resembling a distributed virtual battery.

There are pieces of this already — demand response, dynamic tariffs, OpenADR, smart thermostats, EV managed charging, etc. What I find interesting is that we still don't seem to have a widely adopted flexibility interface that makes arbitrary HVAC/refrigeration equipment a grid resource.

That seems like a fun engineering problem rather than just an electricity-pricing problem. :-)