Australia’s Controlled Loads Are In Hot Water

Australian grids have long run a two-tiered pricing scheme for electricity. In many jurisdictions, regular electricity was charged at a certain rate. Meanwhile, you could get cheaper electricity for certain applications if your home was set up with a “controlled load.” Typically, this involved high energy equipment like pool heaters or hot water heaters.

This scheme has long allowed Australians to save money while keeping their water piping-hot at the same time. However, the electrical grid has changed significantly in the last decade. These controlled loads are starting to look increasingly out of step with what the grid and the consumer needs. What is to be done?

Controlled What Now?

Hot water heaters can draw in excess of 5 kW for hours on end when warming up. Electrical authorities figured that it would be smart to take this huge load on the grid, and shift it to night time, a period of otherwise low demand. Credit: Lewin Day

In Australia, the electricity grid has long relied on a system of “controlled loads” to manage the energy demand from high-consumption appliances, particularly electric hot water heaters. These controlled loads were designed to take advantage of periods when overall electricity demand was lower, traditionally at night. By scheduling energy-intensive activities like heating water during these off-peak hours, utilities could balance the load on the grid and reduce the need for additional power generation capacity during peak times. In turn, households would receive cheaper off-peak electricity rates for energy used by their controlled load.

This system was achieved quite simply. Households would have a special “controlled load” meter in their electrical box. This would measure energy use by the hot water heater, or whatever else the electrical authority had allowed to be hooked up in this manner. The controlled load meter would be set on a timer so the attached circuit would only be powered in the designated off-peak times. Meanwhile, the rest of the home’s electrical circuits would be connected to the main electrical meter which would provide power 24 hours a day.

By and large, this system worked well. However, it did lead to more than a few larger families running out of hot water on the regular. For example, you might have had a 250 liter hot water heater. Hooked up as a controlled load, it would heat up overnight and switch off around 7 AM. Two or three showers later, the hot water heater would have delivered all its hot water, and you’d be stuck without any more until it switched back on at night.

Historically, most electric hot water heaters were set to run during the low-demand night period, typically after 10 PM. Historically, the demand for electricity was low at this time, while peak demand was in the day time. It made sense to take the huge load from everyone’s hot water system, and move all that demand to the otherwise quiet night period. This lowered the daytime peak, reducing demand on the grid, in turn slashing infrastructure and generation costs. It had the effect of keeping the demand curve flatter throughout the whole 24-hour period.

This strategy was particularly effective in a grid predominantly powered by coal-fired power stations, which operated most efficiently when running continuously at a stable output. By shifting the hot water heating load to nighttime, utilities could maintain a more consistent demand for electricity throughout the day and night, reducing the need for sudden increases in generation capacity during peak times.

 

 

Everything Changed

The Australian grid now sees large peaks in solar generation during the day. Credit: APVI.org.au via screenshot

However, the energy landscape in Australia has undergone a significant transformation in recent years. This has been primarily driven by the rapid growth of renewable energy sources, particularly home solar generation. As a result, the dynamics of electricity supply and demand have changed, prompting a reevaluation of the traditional approach to controlled loads.

Renewable energy has completely changed the way supply and demand works in the Australian grid. These days, energy is abundant while the sun is up. During the middle of the day, wholesale energy prices routinely plummet below $0.10 / kWh as the sun bears down on thousands upon thousands of solar panels across the country. Energy becomes incredibly cheap. Meanwhile, at night, energy is now very expensive. The solar panels are all contributing nothing, and it becomes the job of coal and gas generators to carry the majority of the burden. Fossil fuels are increasingly expensive, and spikes in the wholesale price are not uncommon, at times exceeding $10 / kWh.

Solar power generation peaks are now so high that Australian cities often produce more electricity than is needed to meet demand. This excess solar energy has led to periods where electricity prices can be very low, or even negative, due to the abundance of renewable energy on the grid. As a result, there is a growing argument that it now makes more sense to shift controlled loads, such as hot water heaters, to run during the daytime rather than at night.

The rise of home solar generation has created unexpected flow-on effects for Australia’s power grid. Credit: Wayne National Forest, CC BY 2.0

Shifting controlled loads to the daytime would help absorb the surplus solar energy. This would reduce the need for grid authorities to kick renewable generators off the grid in times of excess. It would also help mitigate the so-called “duck curve” effect, where the demand for electricity sharply increases in the late afternoon and early evening as solar generation declines, leading to a steep ramp-up in non-renewable generation. By using excess solar energy to power controlled loads during the day, the overall demand on the grid would be more balanced, and the reliance on fossil fuels during peak times could be reduced.

Implementing this shift would require adjustments to the current tariff structures and perhaps the installation of smart meters capable of dynamically managing when controlled loads are activated based on real-time grid conditions. In a blessed serendipity, some Australian states—like Victoria—have already achieved near-100% penetration of smart meters. Others are still in the process of rollout, aiming for near 100% coverage by 2030. While these changes would involve some initial investment, the long-term benefits, including greater integration of renewable energy, reduced carbon emissions, and potentially lower electricity costs for consumers, make it a compelling option.

Fundamentally, it makes no sense for controlled loads to continue running as they have done for decades. Millions of Australians are now paying to heat their water during higher-demand periods where energy is more expensive. This can be particularly punitive for those on regularly-updated live tariffs that change with the current wholesale energy price. Those customers will sit by, watching cheap solar energy effectively go to waste during a sunny day, before their water heater finally kicks at night when the coal generators are going their hardest.

While the traditional approach to controlled loads in Australia has served the grid well in the past, the rise of renewable energy has changed things. The abundance of solar generation necessitates a rethinking of when these loads are scheduled. By shifting the operation of controlled loads like hot water heaters to the daytime, Australia can make better use of its abundant renewable energy resources, improve grid stability, and move closer to its sustainability goals. It’s a simple idea that makes a lot of sense. Here’s waiting for the broader power authorities to step up and make the change.

125 thoughts on “Australia’s Controlled Loads Are In Hot Water

  1. I’ve lived in places in the US where we had something kind of similar.
    Ours wasn’t a timer in the house though.
    Hah! If we had that here I bet there would be a non-negligible number of people who would try to bypass the timer!

    Ours was a separate utility power line that was cheaper but less constant. It could go off for some number of minutes at a time. I guess when the power company needed to shed load? I don’t know. It could be on and off a bunch but you would never notice with it’s intended purpose because one was unlikely to run out of hot water during the short off times.

    I don’t know if this is still a thing. I’ve moved since I last saw it. Today people might just hook a power wall up to something like that and have cheap, probably illegal electricity.

    1. As an Aussie I need to correct this timer suggestion. If there is a timer used at all it is at the the power generator. The units at the premises are called relays and they are switched on and off by a frequency sense down the power line. These relays can already be bypassed by removing the neutral wire by which the unit receives it’s signal. As an electrician I have had to do this many times when the unit has failed and manually switch the unit back on and notify the power company to come and replace the unit.
      As for wholesale power pricing don’t get me started on this bs. It is an artificially created market to fill the pockets of corporations who have taken over the planet and have more power than the governments who are supposed to serve the people.
      I wonder sometimes if the stupid smart meter emission theories weren’t started by the power companies themselves to distract people from the true nastiness of introduction of peak pricing.
      Australia is abundant in all necesary resources for construction and energy and yet pays the same if not more than other countries less resource rich.

      1. You clearly have little to no understanding of how grid scale generation works.

        The grid needs to have generation and load balanced. It is not as simple as turning a dial on the coal plant right after people turn their AC units on.
        Solar and wind make this MUCH harder.

        What do you think is a better idea?
        Applying the breaks on a wind farm when demand goes down and the wind still blows, causing extra wear to the system?
        Or giving people some extra cheap/free electricity to “create” more load on the grid when you need it?

        1. Of course power plants can’t react instantaneously to demand fluctuations. They do know when to expect changes in consumption however or they wouldn’t have run controlled load tariffs for so long already. But charging more because everybody needs to cook their supper in the evening when they get home from work is not a fair way of handling the problem.

          1. Yes it is, because every time you need to adjust the output of a large and slow generator unit faster than it can respond, you actually have to fire up smaller and faster units that use more expensive fuels less efficiently.

            For example, if you want to adjust your generator down, you start the down ramp hours before the actual event and make up the difference in instantaneous demand using gas turbines or even large diesel generators. This costs more money, so of course you have to pay for it.

            The more the grid demand fluctuates up and down on a short time scale, the smaller the portion of the power that can be supplied by large and cheap generators like CCGT, combined power and heat, or nuclear etc. that simply cannot ramp up and down that fast. The area under the curve that meets the fluctuating demand will be supplied with more expensive sources and/or imports where hydroelectric power is not available.

        2. I saw an article a while ago where a guy hooked the solar panels up to a dc heater element in his hot water tank with a backup thermostaticly controlled AC heater for cloudy days , worked a treat and slashed his electric bill to almost nothing. Seems the simple solution to me , no expensive grid sync inverter, or battery pack, just dump the solar generated electricity into heating a big water heater.

        3. You sound like someone paid off by the power companies?
          We have quantum computing, rockets, nuclear power, x-rays and MRI.. but matching generation against usage is somehow impossible.

      2. +1 on the “not a timer”. Engineer and son of the head protection engineer at the NSW electrical distribution company. The signal is the “deep blue” system. Several different signals are sent out at various times from about 8pm. It’s fun to watch them on cheap halogen lamps. Makes the tin foil fanciers nervous tho.

        1. Yep. The system that has been there for many decades actually allows for over a dozen different codes/groups, where each house’s off-peak (CL) relay would be assigned to a group. My cousin did his thesis on how much better the demand could be managed if those were all used, but at the time everyone was all in the same one. We did advance to “CL1” & “CL2” where CL2 would get a top-up late in the morning but left it there ages ago.
          Now we have “smart meters” which are really just time-of-day categorising meters.

          The whole concept does warrant revisiting. Coal-fired is good for constant power generation. Solar is good for “when the sun shines”. Both are quite different to our consumption (demand) patterns. Anything that helps us time-shift consumption would improve efficiency. e.g. storing heat in water that you eventually want hot. Or, heat-storage space heaters, which heat bricks inside an enclosure, and later run fans over them to heat the air.

          The thing that makes it difficult to align generation with consumption is the market disconnect. There’s a wholesale energy layer. There’s regulations around how the retail bill can be structured. While distribution is a natural monopoly (makes no sense to have competing sets of wires), our retail market is required to be “competitive” within the billing regulations, above a separate wholesale market which has forward and spot pricing.

          1. That’s fine for Joe Bloggs consumer, but you can also have market-rate pricing direct to consumers on an opt-in basis. You advertise that tariff and that’s what they pay.
            We already have a similar tariff available in the UK.
            You have a live display of what the unit price is and do your washing etc when it’s cheap (or, even better, negative).
            Theoretically you could put individual appliances on a smart switch that switched them in under a certain price threshold automatically.

    2. You get a better electricity price if you allow the utility to install a radio controlled disconnector on your central AC. IIRC marketed as ‘Peak corps’

      Faraday cage and ur done.

      1. It’s also fairly common usage in the half of the US that I’ve lived in.

        Depending on what you call “hot”, it does at times add heat to water that’s already hot (hotter than it was when it first went into the tank anyway). If it doesn’t ever do that then there’s either something wrong with it or it’s a demand heater.

        Having said that, once I had thought about it, I just started saying “water heater”.

    1. You can heat 50C water, which is pretty hot, all the way up to 100C. Although you shouldn’t really go beyond 60C because of the risk of severe scalds. And a modern hot water heater won’t heat hot water to anywhere near 100C. So don’t heat hot water with a hot water heater too hot.

      1. That is what thermostatic mixing valves are for. Many solar hot water system of the stored pressure variety can heat the water to 100 degrees and depending on the pressure even slightly higher before the relief valve goes.

        That is why there are laws in place that require a thermostatic mixing valve for any hot water taps in “personal care” spaces (bathrooms) so you can dial your delivered temperature down to 55 degrees.

        In fact with stored pressure style hot water systems you WANT the temp to go above 70 regularly to make sure you are not going to get legionella.

          1. Yep, you need 45 C to kill legionella. 50 C will kill everything else eventually.

            Internally the hot water systems run at 70 C which kills everything pretty much instantly. As @cunning says the thermostatic mixing valve then drops that to 50 C or below (50 C is the maximum allowed at the tap.).

  2. I don’t have much long-term experience with those on demand insta water heaters, but wouldn’t it be a better option? Or is the energy cost something wildly off compared to a traditional tank heater?

    1. No, it wouldn’t! At least not as long as you weren’t always taking a shower or washing your clothes at midday. (Which you’re not.)

      The trick here was always saving up energy in the form of hot water. The problem now is that the times that it’s cheap have phase shifted by 180 degrees, and the regulations need to catch up with reality.

    2. The instant hot water as a concept really only works great for really really tiny spaces, with low occupancy that can’t fit a large enough hot water tank to effectively hold lots of energy from whenever it is most abundant and have enough ready to use whenever you need the hot water.

      So for your tiny homes, caravan, many commercial/industrial places the instant hot water only when you need it can be great – you don’t need much hot water, and you likely have nowhere to store the big box full of it either. For everyone else that actually uses a reasonable amount of hot water being able to heat that big and well insulated tank full up whenever energy is abundant is far superior – yes some energy is lost through the tank insulation, but that can be kept to a very small amount, and losing it when as it the case in Aus now that energy could have come from the Renewables that are being shed in the daytime as there is too much renewable its irrelevent – you have gone from wasting Kw of otherwise effectively free energy to wasting a few watts…

      1. I think you are correct for the most of that. 1. instant and on demand are different concepts.
        2. I have an on demand and love it for endless hot water. 3. I pay less in energy cost for this on demand unit then I do for the older tank type. 4. they have hybrid units that offer the best of both worlds. 4. I live in a not so tiny home and the water pipes are not insulated.

      2. When electric on demand water heaters were introduced 40? Years ago here in northern us the had max delta T of 50 F and winter water temps coming in at 40 F they were a joke snd disapeared quiqkly. New gen ones arived tha supposedly work fine if you have the electrical service for it and can afford the power. People love their gas on demand water heaters though. Giant house, 5 teenagers, no problem even in winter.

    3. I’ve had the gas powered on demand water heater, which turns on after you turn on the hot water and it heats it as you go via gas fire and copper fins on a (what’s the opposite of radiator) water heater. I’ve also had, in the same country, an electric water heater which heats a tank of water and is insulated to stop the loss of heat ( which really just heats upy home constantly) and the cost is about the same, but in the summer you have to block off the areas you don’t use as much because the heat is hard to control despite the insulation.

      I find that both systems cost more or less the same in a situation wherein electricity is more or less always the same price, and environmental temperatures are hotter during the day.

      This article is interesting and applicable to a lot of people, but your milage may vary, so don’t discount others who live somewhere else when they tell you what they pay.

  3. We have a similar tariff in the UK, called Economy 7:
    https://en.wikipedia.org/wiki/Economy_7

    @Panondorf: changing the timer would just mean that you’d get charged a higher rate for electricity used outside the specified off-peak period — the timer is to ensure that appliances that you want to use the cheaper rate are only switched on during the off-peak period.

    1. I’m surprised it hasn’t become as much of an issue in the UK as it’s somewhat similar to Australia.
      If you look at things like Octopus Energy’s “Agile” tariff (hourly rate metering, decided the day before), the overnight rate isn’t massively less than a nominal-daytime rate, just with a massive spike 4-7pm as most people get home and cook dinner.
      The most notable thing is that during the daytime, helped because of our solar production, there are even days on the 12-2pm area where one is paid to use electricity (such is the excess) – can make it a nice money earner to have battery storage charged then and then discharged a couple of hours later when the grid wants it.

  4. I hate to be pedantic (okay, maybe I don’t), but a hot water heater (used 9 times) is an incorrect term. Why would anyone heat water that’s already hot. Let’s go with just “water heater” for the next article.

    1. Uhh, nope? The whole point of this article is about shifting the load of heating water, and when doing that the conventional way you DO heat it while it’s still hot, instead of letting it get cold first before heating it back up again. But when shifting the load, if you don’t have enough stored heat in the water you might end up heating cold water when the power is reapplied. Since it’s built to be a hot water heater instead of an instant water heater, it will take a long time before the water is hot again. If it was an instant water heater, it could and would heat cold water to make it hot.

  5. Welcome to Europe. ;-)

    Don’t control the load, offer the prices and let the individual household pick the energy consumption by it. There are a number of different tariff schemes for that already. Know how they are used by having smart meters.

    Also it is much more efficient to run a hot water heater as soon as possible and not when it is completely cold.

    1. Can you provide a source on that?

      Blackbody radiation goes up with the temperature to the 4th power, so hot objects lose heat significantly faster than cold.

      Engineering often is at odds with pure physics though, so I’m not saying you’re wrong, just interested in knowing why you’re right!

      1. Yeah, that’s been a very persistent myth. That it’s more efficient to keep something warm forever rather than having to heat it all the way up from cold again.

        Being kept hot doesn’t magically “pause” heat loss, it’s still trickling away just as fast as it would if it weren’t being actively warmed by an energy input. In fact the greater the heat gradient, the faster the loss.

        See also: the idea that it’s more fuel-efficient to just leave a motor idling than to start it again. I think that started with some kind of massive WWII tank engines. I don’t believe it was true even for those.

        1. In fact my parents did the experiment with room temperature lowering, and not letting it cool completely down had better results, presumably because the heaters had better efficiency at constant low load than at a short high load.
          For the motors, it is probably not better fuel-wise to let them idle, but for the wear, since on start-up the oil film for lubricating has to be recreated, and the motor has some thermal cycling while switched off and back on. Depending on the machine you can calculate , how long it has to be switched off, so that the fuel cost offset the higher maintainance cost caused by the higher number of machine starts (you will have to determine the effect of the starts on the maintainance cost, though, which may be problematic).
          Same for lighting: if switching lowers the lifetime, it may be better not to switch it off for short times, especially where changing a light bulb is difficult/expensive.

          1. There’s no obvious place that energy going into the heating elements could end up without first passing through the water that surrounds them, so how could the elements ever change their level of efficiency noticeably?

            As for lighting, it’s moot at this point because LEDs can be switched a bajillion times without issue (as in PWM), and so can switching regulators, for obvious reasons. Though a lower constant current beats a PWM’d full current for longevity and efficiency, if we start looking at dimming, and often cheap circuits suck even though their leds are still fine.

          2. In our case it is a oil-fueled heater, and on higher load the exhaust gases are warmer.
            The point with the lighting has always been moot, since it applied only to really hard to reach lamps, but the argument for that special case got stuck in peoples mind as generally applicable (a “runaway argument”?).

          3. so how could the elements ever change their level of efficiency noticeably?

            The wires to the heater have resistance too, and the I2R losses all the way back to the power station are greater when running at higher instant load.

          4. @Dude
            Hence the “noticeably” because the loss should be an insignificant fraction of the total power unless the wiring is poor when the load is purely resistive and well inside the ratings. So in that case their efficiency is fine to consider as invariant, unlike oil, given reasonable conditions.

          5. the loss should be an insignificant fraction

            Normal grid loss is in the 3-8% range. For long distance transmission, it can go as high as 13%.

            Doubling the current quadruples the loss, which is not insignificant.

          6. @Dude: You’re supposed to be showing that a conventional resistive water heater with a tank is more efficient if you keep it hot at all times than if you let it get cold and then re-warm it. If you can’t find a place inside the home at which the efficiency is worse, then you can’t look at the water heater in isolation anymore and just argue based on I squared R. There is little difference between consuming one unit power for one minute or two units power for thirty seconds, because you average out with other people by randomly consuming power on slightly different schedules and so there’s only coarser patterns remaining. One such remaining pattern is people all kicking their heaters on at the same time of day to take advantage of lower rates. If that spikes load enough, that could drop efficiency. More importantly, some severe spike could swing the supply and demand around. But luckily we know that while the ideal time of day is changing, the power companies still find controllable loads to be a benefit if done at the right times, and so any difference in efficiency by the less-random distribution of power draw must not be significant by comparison. If, as a made up example, the loss percentage went from 8 to 8.8 percent during solar hours, but a peaker plant didn’t have to run that evening, that sounds like a good trade.

        2. The engine thing I suspect is just a mis-understanding of a simple fact – that it takes a MUCH richer mixture to start an engine than keep it running. But, starting is only a few seconds of that richer mixture so it never made any sense anyway, people are just bad at thinking about this stuff.

          Same myth with fluorescent tube lights – they took a kick at startup for a second or so and then settled down / warmed up but people would leave them on 24/7 to “save” electricity.

          1. I once worked out the time period that it made sense to leave a CFL bulb switched on, and IIRC at that time it worked out to about 7-8 seconds. So if you left the room for more than ten seconds you saved money by turning the light out.

          2. Electric motors are the same. That’s why clothes washing machines have 2kv starter capacitors. They give an extra kick to get the thing going, then don’t do much at all.

        3. Well, to be fair the way people generally “prove” you shouldn’t idle or that you shouldn’t leave a light bulb turned on is also flawed, even though the conclusion is right on average. They assume that any extra consumption attributed to the stopping and starting will occur immediately.

          It’s entirely possible that if you stopped for a few minutes in the cold, your engine consumes extra fuel for several minutes after being restarted as it tries to get back up to normal operating temperature so that it can enter closed loop. Or, with the tanks, maybe it wouldn’t run properly unless warm. I think they often just ran them for a fraction of the time on a rotation, to keep everything ready to go. Other things that could have an effect, on that topic, even though they don’t directly make the consumption worse than leaving it idling in general: it must recharge the battery from the time when it was off and being drained by all the electrical things that were still running, so the alternator will work a little harder for awhile. Or if it’s hot, it has to work the A/C harder once it starts up again.

          The real cost of such things is usually accelerated wear – which could make your efficiency drop over the long term, or could just represent the need to buy a new item sooner, with a new embodied energy.

          Since water heaters don’t seem like they have the same problems with that, reducing heat loss is great if there’s a period of time you won’t need the storage buffer for. And always assuming it’s not a situation like e.g. you have a heat pump water heater and it’s summertime so even sending fairly small amounts of hot water down the drain is enough to net cool the house despite the heat leaking.

        4. There is a small, and probably insignificant, counterargument. A heater that can heat water rapidly will draw more current than a slow heating unit. The higher current means more losses the house wiring, or more expensive wire to reduce the losses.

      2. Metering our daily consumption dropped by around 1kWh/day since we got rid of the night tariff and went for unsmart instant heating. That’s a large boiler (as they are called here) but not boiling rather going for 40 to 60°C with two to three 10 minute showers per day.

  6. Here in the Netherlands a little bit different scheme has been used for 30+ years. We have had an option to have a “dual meter” installed with two mechanical counters. If you opted for this meter, then there were “cheap” and “expensive” hours. So electricity is always available, and it’s your own responsibility when you switch heavy loads on or off. So if your boiler needs an extra heating cycle during the day because you have visitors, then it’s not a problem.

    You did have to make a lot of use of the electricity of the “cheap” hours to compensate for the “expensive” hours, (which was a higher price then the “averaged” cost.

    I have heard some rumors this is going to change in a few years (may already have changed a few times already). Apparently with the electronic meters, we are going to get a system where it is possible that the cost can change per hour. I do not know how easy it would be to synchronize loads with the changing rates. Here in the Netherlands the government has a reputation for half thought out bad compromises, and still the country seems to do fairly well.

    For myself, it does not matter at all, I just have a rental apartment, no possibility at all for solar cells and don’t use much electricity anyway. (Paying around EUR50 per month for electricity, not much can be gained there one way or the other.

    1. Years ago, we were doing work for the Swamp German power company.

      One of the components had a bug. On a conference call the boss told the Chinese PhD that if he didn’t fix it ASAP he was going to be sent to Amsterdam, to fix it at the client site.

      I asked: ‘What can I break to get sent to Amsterdam? Say the word and it’s as good as broken.’

      I can confirm that swamp Germans do have a sense of humor.
      The client was amused.

  7. 2 or 3 decades ago at a residential curb I found 3 water tight plastic boxes with VHF XTAL controlled radios to drive a big relay. Parts! Someone I knew said they were hooked up to water heaters on customers of a rural electric coop in the area. Tinfoil could have defeated them? The hinged boxes are strong I used one as a speaker box in my slide guitar.
    Solar is big here too so the radio approach could switch on more smartly.

  8. The controlled load meter would be set on a timer so the attached circuit would only be powered in the designated off-peak times.

    Before the invention of the smart grid, we used to have a two-tier pricing system where the electricity company would send an overtone on the grid itself to tell the meter to switch to a different counter and turn on any secondary loads like space or water heaters. It was a simple analog vibrating reed inside the meter, where the presence of the overtone, or a particular combination of tones, would trigger a relay and latch the meter to count discount hours.

    The same system was used to control cut-off switches along the grid to isolate branches, but they tore it down in the late 90’s and replaced it with radio control and simple timers.

    1. Here in the UK the switching signal came over the radio waves, and that’s causing its own problems now as the relevant transmitters are due to be shut down next year, causing millions of meters to suddenly stop working properly. The electricity companies are trying to upgrade households affected, but as that will mean installing the dreaded smart meters, lots of people are trying to hold off until the last minute.

      1. I think the death of the old analog signaling system came from CFLs and other switching mode power supplies simply swamping the signals they were sending, so they could no longer reliably operate it. The other issue was that the signal would pass poorly through transformers, because they would act as chokes for anything higher than 50 Hz, so they had to use a lot of power to modulate the carrier at the source.

        There was a distinct sound to it as well. Fridges and other things with motors and coils would start resonating with a slightly louder buzz after 10 pm.

        1. In fact, you don’t need smart meters nor communication to them. Just measure the frequency and switch the lower price and the load when it is high (== measure the average over the last 24 h and the instantaneous value, high is whenever the last one is higher than the former one. That is self-calibrating and doesn’t even need a timer, only perhaps some hysteresis).
          The transformers blocking the signals may be considered a feature, since it allows to put a signal on single subnets.

          1. It’s not a realiable method, and can have unintended consequences with that sort of dumb frequency followers. It’s too easy to create a chaotic oscillator that way.

            These days with fewer actual spinning generators and large induction motors on the grid, the frequency does not necessarily follow the load at all. If it’s all inverters, then the frequency doesn’t even matter – except when you get conflicts and start running out of sync somewhere.

          2. These dumb frequency followers are exactly what the rotating masses have been in the past. Using the frequency as measure for load balance is not a bug, it’s a feature. Otherwise you would have to implement a second connection to transmit that data, and if it is not extremely fast (which it isn’t, because you have to measure, encode, send, receive, decode and react to changes) you just built your chaotic oscillator. And it has all the digital security issues as cherry on top.

          3. The spinning generators and motors respond more or less linearly to load variations – they’re not on/off like your boiler would be. The step response is what makes the system chaotic, when many devices respond in the same way and cause a ripple in the grid frequency instead of a smooth gradual change.

            Even with the generators providing inertia on the grid, a sudden load drop can and will cause it to trip up, as has happened many times in the past.

          4. You wouldn’t use that on multi-megawatt-units, as you wouldn’t use ripple relays on them. For the small appliances on the W and kW scale the calibration differences of the frequency counters smooth out any steps below what is recognizable on net scale.

            The european grid is designed to ride through any 3000 MW step, so a bunch of badly coordinated 1 kW or even 10 kW steps are very unlikely to do more harm than good.

          5. You wouldn’t use that on multi-megawatt-units

            Take some kilowatt-sized units and multiply that by a few million households, and what do you have?

            The “calibration difference” is moot when there’s a larger frequency ripple on the grid that takes it beyond that “fuzzy” region and causes most of them to switch on or off at the same instant.

            What they would need is a randomized delay or a ramp function that fades the load in and out, but that would require some sort of smart controller rather than just a dumb frequency triggered switch, and so we’re back in the smart grid domain.

          6. The european grid is designed to ride through any 3000 MW step

            If you also add a bunch of frequency sensitive dumb switches on the grid, you can count on having an amplifying effect.

            For example, one million households across the network running 5 kW boilers = 5000 MW. At the sudden loss of 3000 MW of power, the frequency sags hard and all of those boilers turn off. Now you have an excess of 2000 MW on the grid and the frequency starts going up rapidly until all the boilers turn on… flip, flop, flip, flop… with an amplitude of 5000 MW.

          7. So this is what happens when you have a smart grid with its delays. With the frequency counter the boilers switch off fast enough to just stop the sag (an instantaneous power loss doesn’t cause a frequency jump, but a frequency drop with an at first constant df/dt, which then is catched by the control mechanisms within a few seconds).
            Oscillating can not occur, since high power appliances usually don’t like fast power cycles, so after switching on or off you keep that state for a reasonable time. In addition to the calibration differences you can further distribute the reactions by scaling the setpoint with the frequency difference; with this there will be no two appliances switching always at the same time.

          8. Oscillating can not occur, since high power appliances usually don’t like fast power cycles, so after switching on or off you keep that state for a reasonable time.

            It’s a heater – just a bar of metal twisted in a loop – it doesn’t care. It will turn on as fast as you can switch. Again, your switch needs to be smarter than just turning on and off at some fixed frequency limit: it needs delay and hysteresis and fault condition checking…. it’s really akin to a smart meter.

          9. It’s a pretty big relais, and that limits how fast you want to switch. All other things you can get with a CD4040, some CD4029 and CD4008, a control lamp and a 555 (or a really small microcontroller). That is pretty much within hackers reach and far away from the infrastructure you need for accomplishing the same with anything “smart”.
            For starters, don’t switch a load, but an indicator LED (or LED-bar with a CD4051), and shift your tea kettle on-time accordingly. Or just watch it to get a feeling for the frequency shifts. I almost did this, but before finishing the debugging I had to move, and now it is somewhere deep down in a box. I’ll resume when (if) I find it :)

  9. Fossil fuels are “more expensive” only because of government interference in the markets, making them “political” issues to curry favor with certain elements of the population. Command economies have never worked; they suffer from “pilot-induced oscillations”, every time.

    1. If energy extraction companies had to pay market price for most of the land they’re using for coal mining, they would be much “more expensive” than they are now. The IMF says that right now governments around the world are subsidizing fossil fuel exploration and extraction by $7 trillion per year. That’s 7% of the world gross domestic product.
      The free market has very little to do with fossil fuel vs renewable energy prices. Nor does currying favor with certain elements of the population, other than the element that’s pocketing the record profits that non-renewable energy extraction companies are posting.

      1. You should suspect those numbers.

        We’ve seen the lies told.
        Writing off expenses is called a subsidy (but only for energy companies).
        Taxes on gasoline are called subsidies by greenies.
        Claims are repeatedly made that all airports are subsidized. In fact all large airports are cash cows.

      2. “The IMF says that right now governments around the world are subsidizing fossil fuel exploration and extraction by $7 trillion per year. That’s 7% of the world gross domestic product.”

        The IMF? now there is a reputable organization…. Not!
        The sheeple will believe almost anything these days… common sense should tell you this is not a believable number…
        Nuf said…

  10. Experiences here in the Usa –
    I worked on some chiller systems that spent the night running the primary loop water as cold as they could ; routing the typical A/C building load away to a large water tank that could deal with the expansion and making ice all night long. Then during the day the building loop and ice ‘block’ would be connected together (chillers off as much as possible).
    The commercial metering system had on and off peak recording (kW) that rolled on a six month block. If the chiller needed to come on in-the-day the peak for the sixth months would be set and the utility payer would be facing higher bills until it self reset.

    Then in the 90s my house had a FM radio box right above the water heater that would allow the utility to remotely switch off the power. Usually they were good about waiting until right at a city wide peak event to shed those loads. They did not meter it differently rather just gave a rebate every month perpetually after box install.

    I also got involved in a ill-conceived load shedding scheme with some restaurants that tried to knock back the dining and kitchen air conditioning during peak conditions. It made sense to some bean counters but ran afoul of the fact that kitchens and dining rooms need Maximum cooling right at the hottest (building) temps of the day. Competing Interests.

    1. To cool a professional sports venue in Phoenix, they built a similar system. The system (actually owned by the county) was run as a small utility and supplied chilled water to other buildings in the vicinity of the venue. I knew about the system being having been built, as it was touted heavily before construction. I didn’t know about the utility service aspect until a guy who used to manage it told me about it. I found that fact to make the system even more interesting.

      1. One of the systems I didn’t really work on but that pumping company did sell the parts on ; they took that same water tank for ice, and drilled two giant holes on either side. They piped up the loops from a server farm (a gigantic farm). The ice tank was kept full so that if and when the Main utility power failed the cooling loops could get access to a heat dump very quickly. I guess the theory was that without sustained Mains power the Farm’s reserve battery and generators had capacity for Pumping only. They couldn’t do chillers or keep cooling towers running. But with the ice tank enough heat rejection capacity existed to allow the farm to not melt itself.

        1. Ice/water tanks are common parts of large building cooling systems.

          The real trick is getting the ice to self detach from the cooling plates after reaching a certain thickness.

          The cooling loop itself is usually filled with antifreeze.

    2. I wonder if the kitchens aren’t a bit like datacenters… There’s good math to suggest that actively cooling datacenters is generally a waste, and introduces points of failure. The more reasonable path is just mass air ingestion.

      I’ve been in a couple of commercial kitchens that seem to present the same sort of problem. It’s bloody hot, even with the AC blasting. Hotter than it gets outside, baring maybe a few days a year. A whopping massive forced air intake along with boosting the extraction may be more reasonable than air conditioning.

      The hyperscalers are doing this with datacenters. Generally don’t see commercial operations doing the same, as of yet.

    1. Making the government the single point of failure?
      Giving the government control over each house/business?
      Giving the government “Eminent Domain” over power line routes?

      Looking forward to it.

    1. Yes, electricity prices vary a lot worldwide. In some countries it is common to have electric heating (both for water and the whole house) In Norway for example nearly all electricity is generated by hydro-power and electricity is quite cheap over there.

      https://en.wikipedia.org/wiki/Electricity_sector_in_Norway

      Here in The Netherlands the new trend is to reduce gas consumption, and heat pumps for house heating are becoming more common.

      In the place where I live there is city heating. Hot water pipes run underground and come from … far away. I don’t have gas, use electricity to boil water for tea and showers are via a heat exchanger form the hot underground water pipes.

      A few years the city-heating was extended. About 2 kilometers away there is a water purification plant which collects bio-gas. This gas is pumped to a small heat / power combination (about 2TEU in size) and it’s used to generate both electricity and as additional heat source for the city heating.

      Overall, I’m not very happy with the city heating system. I pay around EUR50 per month for the privilege of being connected to the system, while I use around EUR5 per month of heat from the system.

  11. Cogs.

    The “Off peak” controllers ARE not simply timers that turn off at night. They are receivers that listen to codes and control the load.

    Short tones are sent out over the power grid. The receiver decodes these tones. It switches on/off the load on demand.

    This system has been in place since I was old enough to understand these kind of things. At least 1/4 century probably going on 1/2 a century.

    Until a recent meter upgrade for solar, I had a controlled load receiver in my house that was installed in the 70s or 80s

  12. Hmm, lots of confusion in this article..
    There are two types of these things in australia
    1) an off peak timed supply, with a cheaper rate. These are done like the timer mentioned in this article, and normally went to a large tank in the roof that heat at night and be used during the day. You would also often have an override switch to swap it to expensive full power. Growing up our family of 8 had one of these, and you didn’t want to have the last shower..

    2) the control load meter with a cheaper rate. These the energy supplier remote controls and guarantees an minimum amount of hours per day. In my current holiday house that is 18 hours out of 24. I have, for example a hot water heater on it.
    When I started on this plan, over 30 years, I measured how long it turned off.. Over 6 months it did 24/7 ie was never off!
    Measuring it recently it was spot on 18/24 – and it’s controlled in 15 minute chunks..

    They have also started inverting the price curve, to start making power cheaper during the day, and more expensive during the night – a trend that I think is going to massively accelerate given I don’t think we are going to have enough power overnight to meet demand in a few years..

    The day period is 9am to 4pm – and given where my holiday house is in the tropics that means
    a) they are making a lot of money for an hour or two before 9am and after 4pm (we have long sunny days)..
    b) I’ve recently done some numbers, and ignoring solar, taking into account battery life and efficiency etc, and is now probably worth having some batteries, charging them from the grid during the day at the low rate, and using them at night…

    1. Or even just batteries of one sort or another at this point. Cranking coal fires is not necessarily better than dropping a load of dirt down a mine shaft or some other such battery.

  13. We call the times “Peak” and “Off Peak” here in Victoria at least, and I think in the rest of Australia. Some Hot Water Service suppliers (another term to argue over) have already made the change. The Victorian government subsidises the replacement of gas water heaters with heat pump units, and some are designed to take advantage of home solar to heat during the day. Depending on what you pay for, there is an override button or an app which lets you heat outside the set daylight hours.

  14. This is called “demand response” in the U.S. It was relatively common for water heaters back in the day, and is coming around again… especially for air conditioning via smart thermostats. Many utilities have an opt-in program where you can receive a small incentive to let them time shift your AC use to shave the peak.

    And cold water’s really not an issue either. Technology Connections released a video within the last year looking at various forms of thermal storage in the home, and his water heater stayed plenty toasty for many hours.

    1. My power supplier here in Colorado wants me to get a smart thermostat, so they can turn off my A/C compressor at times. I am actually okay with turning off the compressor, but I don’t want to put a thermostat inside my home network that talks so some big vendor’s insufficiently secure servers. It would be too much like leaving the digital front door open.

  15. The problem is that the tariffs now need to be more variable so that they just guarantee an amount of power each day and charge less for lower tier tariffs. It’s done in some places in Australia with complex feed in contracts where incoming and outgoing power is charged at variable rates so your system will bring in power only when it’s cheaper and supply power back into the grid when it’s more profitable.

  16. Lewin has simplified the situation down a bit to make it easier to digest; for example, our pool pump in on a separate controlled load (CL) tariff to the hot water system’s. Other people have noted some of the implementation (all provider-side control, not at the meter) differences.

    ‘”In a blessed serendipity, some Australian states—like Victoria—have already achieved near-100% penetration of smart meters. ”

    A lot of the article comes across as shilling for the power companies and governing bodies, but this is where the rubber hits the road for the consumers. The anecdotal evidence out of Victoria appears to be that consumers have had increased power bills since the installation of the smart meters without change to usage profiles. Here in Queensland I live in a suburb that has had the roll-out completed, and locally the anecdotal evidence is also that bills have increased since the smart meter install, again comparing periods of equivalent usage profiles.

    There’s a lot of regrettable policy and seeming corruption/collusion that’s gotten us to where we are, and keeping us from getting to anywhere reasonable; I imagine that’s true of most places.

  17. Just another example of Load shifting with a twist here in the USA.

    Similar to many other examples, the electric utility offers a reduced rate for customers that agree to load shifting, the difference here is the prevalence of Air Conditioners.

    The twist is that the Utility would send a signal to the AC units to turn off the High-Power AC Condenser for 15 minutes every hour while the low-power House circulation fan continues to run.

    By cycling through 4 different sets of targeted AC’s, the result is only a marginal increase in House temperatures while reducing the heavy AC loads by 25% during peak times.

    I wonder if Australia could do the same thing with their Hot water heaters so that it wouldn’t be just a binary on-or-off situation.

  18. The whole ‘heat water at night’ system was not designed to flatten the load curve, but to use up surplus electricity at night, since the coal fired power stations could only be ramped down to a specified minimum supply

    1. I believe this is the correct situation. Flattening the load is just a side benefit. They charge money for electricity and make a profit. They don’t really care that much if you pay more or waste electricity, it’s just more profit. We’re definitely overthinking this on the motivation aspect. If I were a rich industrialist with a coal powered generator, I’d say do whatever to maximize profit.

      I’m more of a democratic socialist and hacker, so I imagine better, but I don’t pretend that others are the same. They’re typically much more self interested.

      1. Sigh, really?

        It was both. Excess night power used, daytime load reduced, plus customers got a lower rate by switching. There is also the “off peak” system where you are charged a lower rate at night so you run your cryptominers then. (Many places do that.)

        Why would the evil capitalists bother offering a discounted night rate if they could simply charge whatever they liked whenever they liked?

        Most of the excess power from the coal stations at night goes to pumping water back uphill for hydro use, where when used during the day the evil capitalists can charge you the full rate.

  19. also, Mercer (above) is not correct, It is definitely a timer switch (mechanical at that). My parents still have the original mechanical timer in their switchboard, and the (now unused) electric hot water service. The timer was sealed with a copper wire and a lead seal. If you ran out of water, and you were prepared to put up with some fuss, you would cut the seal, and rotate the timer to turn on the heater, then reset it to the correct time a few hours later. There were two ‘meters’ to record energy usage, the hot water one being charged at a different rate.

  20. There are some smart tariffs in the UK that just pass on the half hour wholesale prices to the consumer and then its upto them to optimise accordingly https://octopus.energy/smart/agile/
    I have a little Python script that looks for the cheapest 8 slots a day and puts the hot water heater on then, works well so far and no external authority controlling things.

  21. From someone who has lived and installed many domestic water heaters in Australia, here’s a couple of technical clarifications.

    The control from the utility is not timer based, it is ripple control relays. Pics and info here – https://en.wikipedia.org/wiki/Load_management#Australia_and_New_Zealand
    Most homes have a “boost” switch installed near the heater or near the main panel which overrides the utility load control, for when you run out or know you are going to need more than usual
    This can be taken advantage of using solar hot water diverters such as iBoost, Eddi, etc which override the ripple control and divert solar production into the hot water rather than export it
    The cost difference for what is usually the biggest single consumer of energy in the house can be a 1:3 ratio of night to daytime cost

  22. We used to have a bed side clock radio – when the ripple (frequency) kicked in down the link to turn the offpeak switching in, the clock radio for a few seconds would race ahead in time. 50hz –> to whatever high freq the power companies sent down the line.

    Not the most ideal method of keeping time by the designers of this clock radio.

  23. Serious question

    Is it too expensive to install alternative heating methods and have them plumbed into the hot water reservoir? Even in places as far north as the Olympic Peninsula in Washington state solar hot water heaters work in the winter right up to the point where the snow starts falling. In warmer areas I’ve seen cut off and recirculation valves used when the water got too hot.

    Alternatively there are systems available where the “hot side” of an air conditioning loop can be used to keep a hot water tank warm.

    1. The problem might be the integration of the alternative heating into the old control system, which may be not capable of managing the new heat source, or get confused by the additional heat (not that this keeps happening, but you have to make sure it doesn’t happen).
      For the air conditioning, it has better efficiency if the “hot” side is as cold as possible, so depending of the amount of heat transferred it may be better to release the AC heat at as low as possible temperatures and generate the high hot water temperature separately.

      If everything works out, this is a good way to go.

  24. 250l for 2 showers?
    Who is using 125 litres of hot water per shower?

    I can shower quite effectively with 2 litres, 1l to wet down, then soap up (shower off), then 1l to rinse.
    If I include running it to get warm, that’s 4l tops.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.