4 ms·

Great examples. These measures in the U.K. were put in place to shift demand due to constraints in total system capacity - total generation, total transmission

Great examples. These measures in the U.K. were put in place to shift demand due to constraints in total system capacity - total generation, total transmission, total distribution capacity. Eg a substation transformer couldn’t handle normal daily loads plus heating, so heating load is shifted to night. What happens when power at night becomes more expensive due to increases demand such as car charging or low supply such as high renewable penetration and no wind? It’s all about having the firm capacity to deliver power for hours on end. So the $/mwh of wind or solar power has to go up because grids are going to say you can only have more if you can supply firm capacity. This is already happening in Hawaii. I implemented hot water tank load switching through smart meters on at a town with its own power plant of only 850 kw so they wouldn’t run out of power at dinner time when the river was low (and we didn’t have to fire up the diesel either)

tomatocracy 7y ago
Yes, there will always be baseload demand. But I think with the right setup more demand matching is possible for domestic and commercial buildings with relatively little effort. One way this could work is that you have eg a smart heating thermostat which plans which hours of the day/night to turn heating on and then buys power to cover that in a day-ahead market, with the aim of minimising cost on that basis. Wind and solar power are relatively preditable a day out so you wouldn't be paying too much to power traders for providing day-ahead liquidity vs on-demand pricing. Likewise, EVs all end up parked somewhere during the day - the challenge is ensuring that as many of those places as possible (car parks, workplaces, etc) have charging points too; you could do something very similar there with an algorithm deciding whether to charge based on a spot price and how much is in the battery (say, owner programs car to top up to 65% and then only continue charging if power is cheap).
- generatorguy 7y ago
  How much does it cost to implement and run all the demand matching, how much money does it save the loads who are willing to participate, and how much load can reliably be controlled? The marriage of electrical system reliability to real time control of tens of thousands loads over internet or cellular would be an unholy one. This will limit acceptance as the cost of outages immediately outstrips savings. So I’m back to grids requiring solar and wind to come with energy storage to firm them up. Which is a perfectly fine solution and increasingly viable as the cost per MWH of storage comes down.
  - tomatocracy 7y ago
    I'm not talking about real-time demand matching/load shedding here as a consumer solution - as I mentioned, I think matching on a day-ahead (and perhaps also eg 6-hour ahead) basis could well be a lot better than we have now because wind and solar generation is quite predictable on a day-ahead basis, as are outside temperatures (so eg for space heating energy consumption may be relatively well predictable too). The beauty of this type of approach is that it can be done by incrementally building on things which already exist. There's already an electricity market, including day-ahead pricing in most pools. Professional power traders already trade these markets. Smart meters already exist and are capable of dealing with pool prices not just flat rates. Smart thermostats already exist. EVs already have the capability to stop/start charging based on more than just whether they're plugged in or not. Even better, most of these already have standardised APIs. What's missing is the software stack which ties all of these together - and that can be done incrementally, device by device (on the load side).
    - generatorguy 7y ago
      Thanks for the discussion. The information we are missing is what the percentage of total load can or would be willing to be shifted, and how strong the incentive (savings) is for the load to allow themselves to be shifted versus their inconvenience.
      - fulafel 7y ago
        We know that the answer is 100% as the price has no upper limit, no?

tomatocracy 7y ago

Yes, there will always be baseload demand. But I think with the right setup more demand matching is possible for domestic and commercial buildings with relatively little effort. One way this could work is that you have eg a smart heating thermostat which plans which hours of the day/night to turn heating on and then buys power to cover that in a day-ahead market, with the aim of minimising cost on that basis. Wind and solar power are relatively preditable a day out so you wouldn't be paying too much to power traders for providing day-ahead liquidity vs on-demand pricing. Likewise, EVs all end up parked somewhere during the day - the challenge is ensuring that as many of those places as possible (car parks, workplaces, etc) have charging points too; you could do something very similar there with an algorithm deciding whether to charge based on a spot price and how much is in the battery (say, owner programs car to top up to 65% and then only continue charging if power is cheap).

generatorguy 7y ago
How much does it cost to implement and run all the demand matching, how much money does it save the loads who are willing to participate, and how much load can reliably be controlled? The marriage of electrical system reliability to real time control of tens of thousands loads over internet or cellular would be an unholy one. This will limit acceptance as the cost of outages immediately outstrips savings. So I’m back to grids requiring solar and wind to come with energy storage to firm them up. Which is a perfectly fine solution and increasingly viable as the cost per MWH of storage comes down.
- tomatocracy 7y ago
  I'm not talking about real-time demand matching/load shedding here as a consumer solution - as I mentioned, I think matching on a day-ahead (and perhaps also eg 6-hour ahead) basis could well be a lot better than we have now because wind and solar generation is quite predictable on a day-ahead basis, as are outside temperatures (so eg for space heating energy consumption may be relatively well predictable too). The beauty of this type of approach is that it can be done by incrementally building on things which already exist. There's already an electricity market, including day-ahead pricing in most pools. Professional power traders already trade these markets. Smart meters already exist and are capable of dealing with pool prices not just flat rates. Smart thermostats already exist. EVs already have the capability to stop/start charging based on more than just whether they're plugged in or not. Even better, most of these already have standardised APIs. What's missing is the software stack which ties all of these together - and that can be done incrementally, device by device (on the load side).
  - generatorguy 7y ago
    Thanks for the discussion. The information we are missing is what the percentage of total load can or would be willing to be shifted, and how strong the incentive (savings) is for the load to allow themselves to be shifted versus their inconvenience.
    - fulafel 7y ago
      We know that the answer is 100% as the price has no upper limit, no?