Octopus agile customers: how do you want to control your systems?

One of the reasons we decided to go down the Home Assistant route is that it was flexible enough to express a wide range of peoples control preferences in terms of how people want their system to work. This is generally not possible with other systems which tend to have strong opinions about how you should/shouldnt control your systems. We wanted to do something different and provide something that was capable of doing this (whilst accepting the higher complexity this involves).

This is particularly important where there are multiple different energy smart appliances (e.g. EV charger, diverter, battery etc.) and you are on octopus agile where prices change every 30 minutes every day so you really need to automate things to get the most out of it.

So in this spirit - how do you want to control your system? Is it all about price? How do you want to prioritise different appliances (e.g. should your EV go before your diverter)? How clever do you want the scheduling to be - would you trust us to do it all for you - or is it important that you understand exactly what it is doing and why?


A further bit of context to this for this who are interested… at the moment ~95% of systems will generally be controlled in a fairly dumb ‘load following’ way. For example, if you have a battery it will just monitor your grid feed and charge/discharge to zero this out. And in many cases if you have more than one appliance this will be independently doing the same thing (and then they will be ‘racing’ each other).

But this is quite a dumb way of charging a battery (or anything else). If you have a good solar forecast and some idea of what house load will be (or you can even influence other loads) you can operate the system in a more optimal way (economically speaking).

So this type of control is a possibility in the future. The PowerShaper system has your smart meter data (which could be used to try to estimate your demand to some level) and we are looking at adding solar forecasts. But its complicated and a bit opaque to end users. Would you prefer to set preferences yourself? Or is some hybrid of this and some other simpler control better? I dont think anyone has an answer to this atm and its will partly be up to consumers to decide where this goes.

This is a complex question and for many people is very likely to evolve with the tech that people have available.
In our case we have 2 main items that could / can come under control of Octopus Agile namely: OpenEVSE and Sunamp (Shelly). At some point we may have an ASHP and if possible we would have this under HA / Agile automation.
For the EVSE at the moment I am using the onboard timer functions under OpenEVSE to set the EV to charge between 0130 to 0330hrs. This usually coincides with cheaper Agile pricing and in any case we are currently doing soooo little driving that the car doesn’t often need charging during lock-down.
In terms of our Sunamp it would be excellent to have really smart control of this as this is our household device that will have 1st call on our PV generation. So whilst we would want to top up the Sunamp from cheap Agile elec this would depend on how much PV is available over the next day or so.
With an ASHP we would probably have an automation based on internal (habitable room) temperature and aiming to avoid using peak priced electriciity, so probably boosting internal temps around 0500-0600hrs and around 1430-1600hrs to try an minimise elec use for a couple of hours after those times. But if the habitable room temp dipped below say 18.5C bring the ASHP back into operation.

Type of load Type of Agile or other smart tariff optimisation PV optimisation Params worth including
EV / EVSE Mostly night-time charging. Probably mostly need to add ~40-80miles range per session (ie 10-20kWh) possibly ever day possibly 1/week. Occassional need / desire to divert some PV into an EV but this probably only applies to larger PV arays and from April to Sept Agile tariff (cheapet periods), PV optimisation as 2ndary info
Home Battery Need to top up battery on cheapest Agile rate but with a forecast element to ensure that spare charge capacity available for PV (PV kWh can be considered to cost Zero pence) Priority to the battery (over most other domestic loads) is likely to be wanted by most users. This will be complicated if there is also a Sunamp or HW cylinder with PV divert function. PV optimistion given similar weight to Agile tariff (cheapest periods)
Hot Water Cylinder with PV divert Need to top up battery on cheapest Agile rate but with a forecast element to ensure that spare charge capacity available for PV (PV kWh can be considered to cost Zero pence) Metering the level of HW will be slightly less precise than for a home battery but there is likely to be a priority to ensure that there is always a plentiful supply of HW. PV optimistion given similar weight to Agile tariff (cheapest periods)
Electric heating systems The main aim will be to avoid or minimise using elec from peak rate times (typically 0800-0900 and 1600-1900hrs). This may be achievable by slightly boosting temperature in the house between 0500-0700hrs and 1000-1530hrs. The extent to which this is possible will depend on heating schedule, occupancy levels and solar gains There is only likely to be modest benefit in PV optimisation with ASHP systems, the heat demand will inevitably be highest when it is cold and cloudy. It would only be if a HP system has a fairly significant buffer tank / thermal store that PV optimisation will be really available as an option. Agile tariff (avoid expensive periods), some solar forecasting in order to calculate whether a heating boost before sunrise is worthwhile

Thanks dom this is really interesting and helpful.

@ben I generally agree with @zapaman who has provided a very comprehensive list of wants. Ideally, if you can set up a number of automations that can be customised by individuals that would be great.

Another project would be setting up an open source V2G solution - but this would involve getting the right type of hardware that has also met all the approvals needed for a DNO to approve a G98 or probably a G99 application. Perhaps there could be a way to work with Octopus Energy as they have just announced their London-based research facility will be known as the Octopus Centre for Net Zero (OCNZ) that will create models and policy recommendations for potential paths to a green energy future See Octopus Energy launches London research facility - reNews - Renewable Energy News As we have discussed before there really is a need have an interoperable V2G solution that does not lock in consumers to one vendor (supplier).

This is great Dom. For the optimisation of solar PV divert to water vs cheap night time rates, an integration with Solcast would be good.

Hi @ben @mattfranklin would it be possible to add the Octoblock integration to the HEMS?


I don’t have this tech yet, but in my position (lucky enough that minimising energy bills is not essential), my requirement would be very simple: to minimise my carbon footprint, whilst meeting certain minimum requirements, namely, my car always having at least 20 miles of range in it, house temp not dropping below certain temps, there always being enough hot water for a bath in the evening.

Sorry for the delay! @peter or @Jess_Fairbairn might be able to provide more help on this one, since they have been working on the Octopus Agile integration that they have made available in the HEMS.

Are there specific features you need that aren’t in the scheduler integration that they had previously worked on? That might help to see if there is a way that we could build those features in.

1 Like

Agree with this, trying to do the battery part myself currently, but Tesla Powerwall is hard to control outside of the app… as for solar forcasting I use a Weather underground IFTTT to turn on a virtual switch if its going to be sunny the next day, so as to not charge too much overnight. Which of course brings me back to the powerwall control issues…

Hi it seems like the Scheduler is a simple timer feature. What would be ideal would be a feature that works out (say) the cheapest 4hrs average Agile price over the next day. You could then set a trigger whereby a device would come on if the “current price” is less than this 4hrs average price. In other words it would automate finding the lowest price periods over the next 24hrs. You could adjust the 4hrs to other longer / shorter averages so as to trigger a device for longer / shorter durations. Obvs these periods might not be in 1 block, ie you might have 5no. 1/2hr blocks around 2am and another 3no 1/2hr blocks around 1pm and each day would be different depending on the average price block.

Hi Dom, @Jess_Fairbairn and I are working on improvements to the Octopus Agile integration we are providing, and it should help for this kind of use case.
Our issue at the moment is working out how to get the user configuration working without people accessing the yaml configuration.
Home Assistant development is tricky, quite different from our normal “web development” and has been a bit of a moving target for the last while.
If anyone in the community has python skills and would like to help out, please let us know!

1 Like

Hello Techy Carbon Co-op members, I am wondering about Octopus Agile, and appreciated zapaman’s analysis.
But is there a guide for beginners on all this please? I personally don’t want my life dominated by the changing price of electricity by the hour, but interested in managing demand so systems are more effective/efficient, but would want them fairer too. Also concerns about Octopus’ ethics???
For the record I have solar pvs, a battery, hot water cylinder with pv divert and ASHP.
Not sure it would be a good idea to turn up the thermostat each day then turn it down again.
Anyway interested in how all this is developing, and how new, less techy people can benefit from all your valuable work.
Thanks, Phoebe


Hi @Phoebe_Spence not everyone will be able or happy to deal with fluctuating electricity prices. Nevertheless this type of “agile” tariff is likely to be an increasing feature of an energy system dominated by renewables. Up until now the electricity system has been designed around massive fossil fuel and nuclear power plants and the generators have made sure that electricity is made available to consumers when they demand it. So you and I have been able to turn on appliances without any consideration of the available power on the system; we just turn-ON and expect power to be delivered.
As renewables increase in the mix then there are frequently times when there is too much wind or solar, this means that the power either gets curtailed (wasted) or some users can be encouraged to use more power. Conversely we it is cloudy and still then there will be a lack of renewables on the grid and the only way we can cope at the moment is by turning up the gas fired power stations or “ask” electricity consumers to turn-OFF as many appliances as possible / rely on stored energy.
The best available signal to electricity users at the moment around these types of issues is generally the Agile tariff and some grid carbon intensity apps.
The key to making this stuff work is to make it easy, no one wants to be faffing around with tracking prices and carbon intensity 24/7, so automated systems connected to ASHPs, storage heaters, HW cylinders and home batteries will be key to making this easier.

Hi Phoebe - could you elabourate a bit more about your concerns Re Octopus’ ethics? I’d be interested to know what this issues might be? From a Customer Service point of view they can’t be beaten in my view.

Thank you zapaman, yes fully support this, consumers can help, but schemes using excess energy to pump water to store for future hydro use would be more to scale, plus working with large users of electricity to regulate demand as I believe already happens.
My concern from Ethical consumer: Octopus is a major investor in renewables, and it does own a few of its own. But it is also 20% owned by Origin Energy, which is the largest owner of natural gas-fired power stations in Australia, and we could find no commitment not to build more fossil fuelled plant.
Also I note Origin have encouraged Octopus to move into the Japanese market - the emphasis on ever more growth, no mention of addressing the climate emergency. Quite a business/market approach.
But I read that Octopus Renewables are certainly active in renewable energy, and working with local community energy projects is encouraging. So I personally am fairly reassured, but doubtless more info to come to light.
I’m appreciating these exchanges, helping to clarify and inform my decision making - thank you all

Pumped storage cannot be expected to deal with the scale of change needed. Increasing pumped storage in the UK to the scale needed would involve wholesale filling up of high river valleys. It would involve a great deal of environmental damage. Some modest increases in pumped storage/deferred hydro will be desirable as will some large battery storage projects. Other centralised approaches such as interconnectors to Norway (ie using their hydro etc) will also help.
However, demand side response by consumers is likely to be a key component of the mix. We have to move away from the thought that everything will be “done for us” and will need to be become active participants in the energy system. Partly by energy efficiency measures to reduce overall demand but also by considering (in an automated manner) at what time of day we use energy.

small scale storage for all!

Not sure where to post this, but this looks like a new interesting competitor to Octopus Agile.
first consumer tariff launching nationally based on NationalGrid ESO carbon intensity api. So basically an Agile tariff but with additional discount for low carbon periods and penalties for higher ones.