In a way the BSUoS costs are possibly accounted for, to some degree or other, in the formula that used to determine the Agile price from the Wholesale price with the 2.1 factor for my area. It varies from area to area. That avoids the being caught out with unexpected costs although the situation is slightly perverse because when the renewables are highest, the BSUoS costs are generally highest but this is the time when the Wholesale rate can be zero so the 2.1 factor doesn't apply.

Also, when the Wholesale rate is highest we tend to avoid using power then so it doesn't apply so much there either. If I had a battery I could zero my use at the times of the highest wholesale rates (plus the 12p DUos costs) as well but it doesn't pay for me to do that. I would only save about £80 a year although I would only need a 2kWh battery system at the very most and more likely 1kWh to cover 80% of my peak use.

View attachment 215559

The agile formula is quite simple, for obvious reasons, and doesn't really reflect the cost of delivery in a simple 2.1 multiplier. The peak surcharge does some of this, but for many reasons is well below the increase in costs. There are some regions where the DNO charges for the red peak are >20p/kWh higher than they are in the amber periods. In fact I don't know, off hand, any region where that difference is much less than 12p/kWh. There is a lot of cross-subsidy within Agile, and Octopus made a lot of money off of it when wholesale prices were high. That's fine, in and of itself, but Tracker's formula is much better anchored to actual costs (though you can 'abuse' that too).

For example if Electricity can be purchased wholesale for £10/MWh (1p/kWh), then Agile charges you 2p. Octopus are on the hook for £15.4/MWh (1.54p/kWh) in BSUoS charges from Oct and about £10.5/MWh now. They also must pay the CfD leavy which is about £11/MWh, and the Capacity Market charge (~£10/MWh if I remember correctly). If it is overnight the DUoS charges will be quite low, but in the Amber period they are often around £20-£30/MWh. So that means, before they touch anything else they are paying £40/MWh (4p/kWh) just to supply you. Of course later in the day, when wholesale prices are £80/MWh, they get to charge you 16.8p/kWh, but only have to pay 13.6p/kWh. In the Red peak if prices are £120/MWh, Octopus charge 37.3p/kWh, but will be paying between 32 and 39p/kWh.

The issue with BSUoS charges is that something like 70% of balancing costs are in thermal constraints management (I think it is closer to 80% in actual payments to generators). So these come in when prices are comparatively low. Overall it is quite cheap on a /MWh basis, at least compared to the costs on cold calm January days, but there are a lot. Since negative prices are almost all a function of market design 'failures', then using electricity south of SSE-NS, SCOTEX, or SSHARN3 boundaries marginally adds to these costs. Which are then share out evenly across all rate-payers. Hence the cross subsidy. Of course Agile is a tiny fraction of this demand, most of the 'imablance' is due to the way interconnector users buy power, and make a lot of money off of it. If you live north of SCOTEX use all the power you can in these situations, it will help reduce balancing costs in the mid-term.

I’ll stick with not squandering resources but I’ll use the tariff to my advantage and if Elexon/NESO are unable to improve their wind generation or solar predictions then it is up to them to deal with it as balancing mechanism costs. As always, a system always tends towards a minimum energy or in this case cost. I’m sure it is not beyond the ability of the NESO to find the lowest cost system although that will always be moving goalposts.

The primary problem isn't on Elexon, NESO, or OFGEM can fix. It is set by DENSZ (and its predecessors). The UK wholesale market is structured in such a way as it ignores transmission capacity in its design. It wasn't a problem when the bulk of the generation was in England, and a lot in the South, but it is now. Essentially, wind providers, commercial or CfD participants, can offer power very cheaply onto the grid based on the size of their immediate connection. The big consumers can then buy that power quite cheaply (looking at you inteconnector users) and sell it overseas for a tidy profit. NESO must then replace that power (and compensate the curtailed generator for their 'costs') to ensure that the power is delivered. That power is both more expensive, and almost always higher in CO2 intensity. Note: in some cases the curtailed generator has to pay NESO as there costs might go down. Keep in mind they still get paid for all the power they sold. Curtailing the most recent, most efficient, and lowest strike price CfD wind farms is often the cheapest way to go, as NESO's system goes some way to paying the lost CfD revenue (as CfDs are based on metered delivery and not sold power).

Zonal wholesale markets would have disrupted the inteconnector users profit/subsidy stream, and 'broken' a lot of the business cases for recent wind farms. Which are predicated on selling a given amount of power. It isn't the lost CfD subsidy, but that compounded with the lost sales revenue*. This was the bit that DESNZ found 'intractable' when looking at moving from the current market to the zonal one.

* What's interesting is that the newest wind farms in general would have been fine as prices would spend less time below zero, so there power would be eligible for CfD payments more often. Its the slightly older ones that would have found it hard to sell into that market.

 

The agile formula is quite simple, for obvious reasons, and doesn't really reflect the cost of delivery in a simple 2.1 multiplier. The peak surcharge does some of this, but for many reasons is well below the increase in costs. There are some regions where the DNO charges for the red peak are >20p/kWh higher than they are in the amber periods. In fact I don't know, off hand, any region where that difference is much less than 12p/kWh. There is a lot of cross-subsidy within Agile, and Octopus made a lot of money off of it when wholesale prices were high. That's fine, in and of itself, but Tracker's formula is much better anchored to actual costs (though you can 'abuse' that too).

For example if Electricity can be purchased wholesale for £10/MWh (1p/kWh), then Agile charges you 2p. Octopus are on the hook for £15.4/MWh (1.54p/kWh) in BSUoS charges from Oct and about £10.5/MWh now. They also must pay the CfD leavy which is about £11/MWh, and the Capacity Market charge (~£10/MWh if I remember correctly). If it is overnight the DUoS charges will be quite low, but in the Amber period they are often around £20-£30/MWh. So that means, before they touch anything else they are paying £40/MWh (4p/kWh) just to supply you. Of course later in the day, when wholesale prices are £80/MWh, they get to charge you 16.8p/kWh, but only have to pay 13.6p/kWh. In the Red peak if prices are £120/MWh, Octopus charge 37.3p/kWh, but will be paying between 32 and 39p/kWh.

The issue with BSUoS charges is that something like 70% of balancing costs are in thermal constraints management (I think it is closer to 80% in actual payments to generators). So these come in when prices are comparatively low. Overall it is quite cheap on a /MWh basis, at least compared to the costs on cold calm January days, but there are a lot. Since negative prices are almost all a function of market design 'failures', then using electricity south of SSE-NS, SCOTEX, or SSHARN3 boundaries marginally adds to these costs. Which are then share out evenly across all rate-payers. Hence the cross subsidy. Of course Agile is a tiny fraction of this demand, most of the 'imablance' is due to the way interconnector users buy power, and make a lot of money off of it. If you live north of SCOTEX use all the power you can in these situations, it will help reduce balancing costs in the mid-term.



The primary problem isn't on Elexon, NESO, or OFGEM can fix. It is set by DENSZ (and its predecessors). The UK wholesale market is structured in such a way as it ignores transmission capacity in its design. It wasn't a problem when the bulk of the generation was in England, and a lot in the South, but it is now. Essentially, wind providers, commercial or CfD participants, can offer power very cheaply onto the grid based on the size of their immediate connection. The big consumers can then buy that power quite cheaply (looking at you inteconnector users) and sell it overseas for a tidy profit. NESO must then replace that power (and compensate the curtailed generator for their 'costs') to ensure that the power is delivered. That power is both more expensive, and almost always higher in CO2 intensity. Note: in some cases the curtailed generator has to pay NESO as there costs might go down. Keep in mind they still get paid for all the power they sold. Curtailing the most recent, most efficient, and lowest strike price CfD wind farms is often the cheapest way to go, as NESO's system goes some way to paying the lost CfD revenue (as CfDs are based on metered delivery and not sold power).

Zonal wholesale markets would have disrupted the inteconnector users profit/subsidy stream, and 'broken' a lot of the business cases for recent wind farms. Which are predicated on selling a given amount of power. It isn't the lost CfD subsidy, but that compounded with the lost sales revenue*. This was the bit that DESNZ found 'intractable' when looking at moving from the current market to the zonal one.

* What's interesting is that the newest wind farms in general would have been fine as prices would spend less time below zero, so there power would be eligible for CfD payments more often. Its the slightly older ones that would have found it hard to sell into that market.

I really appreciate the time you take to present the detailed information on how we got here. I come from another well established industry that had effectively tied itself in knots with complexity and the public presentation of that effectively looks like a Brian Rix farce if you are old enough to know what one of those is.

To be honest when I first dowloaded the historic Agile data going back as far as 2018 there was a gob-smacking moment when I realised you could get paid to use electricity. In fact I was able to show that with a big enough battery you can harvest all the negative prices and low cost units up to about 2.4p/kWh and pay precisely zero for a whole year's worth of energy including 11,000 miles in an EV. Of course that is totally impractical (~8MWh battery - someone might want to talk to me about operating that facility on a housing estate) but the fact that this is possible speaks volumes about how broken the system is in places. With each complication comes a new set of loopholes and a licence to print money and Net Zero is playing a part here. I operate ethically in that I don't artificially create a demand when I don't need to but I do use the contract to my advantage otherwise.

I could say that I'm entitled to take advantage where I can because the Agile contract (the formula) says nothing about prices being less than 100p / unit. Perhaps that is the 'reward' for opting to take the risk of wholesale derived prices where you have no prior knowledge of what those might be. With prices that have ranged from -19p to 100p you could say that is highly risky but in practice the average is on a par with discount EV rates with none of the complexity, rules or smart equipment. As far as I'm aware Octopus don't even have to buy any energy on the futures market for me as per Ofgem rules (following Bulb's demise) unless it's to cover the possibility of me switching to another tariff at short notice.

 

I guess the other question would be - as an end consumer, would there have been any way for me to figure out on Monday morning, that actually the morning and not the afternoon was the best time to use power, in spite of wholesale pricing?

In the strict sense, yes. However, it isn't really practical unless you want to spend a lot of effort on it, and react quickly. NESO publish constraint limits and expected flows the day ahead, so you can see which 1/2-hour periods will have higher curtailment for your region. The carbon intensity of the replacement is a bit harder. Then you can also use the in-day market to see what power is selling for on the approach to the gate – that closes 1-hour before the settlement period of interest. That will tell you what the expectation of the additional power that something like Agile engenders will cost; though it is still subject to the constraint problem.

As a consumer I have to go with the price signals that I'm given: we can't go around second guessing pricing strategies and the real fundamentals. If the specialists can't get it right I'm not going to research the issue for optimal system benefits. I did vary my usage using the Greener Days information when I could, but this discussion just reflects a very niche and abstruse commentary on a rapidly changing situation and doesn't reflect the real retail customer experience.

No reason not to, but you also don't want too many other people jumping on these products, as they only work because of cross subsidy. I am not opposed to 1/2-hourly pricing, but the current Agile formula doesn't scale well, especially at lower wholesale prices. Without temporal or regional cross-subsidy and depending on region Agile should never go below something like wholesale, in p/kWh, +6p/kWh overnight, +8p-10p/kWh in the shoulders and+20p-35p/ during the red periods. Of course some of the temporal cross subsidy will be internal to a single customer, but a fair amount of it is between customers. All of the regional cross-subsidies are between customers.

 

In the strict sense, yes. However, it isn't really practical unless you want to spend a lot of effort on it, and react quickly. NESO publish constraint limits and expected flows the day ahead, so you can see which 1/2-hour periods will have higher curtailment for your region. The carbon intensity of the replacement is a bit harder. Then you can also use the in-day market to see what power is selling for on the approach to the gate – that closes 1-hour before the settlement period of interest. That will tell you what the expectation of the additional power that something like Agile engenders will cost; though it is still subject to the constraint problem.




No reason not to, but you also don't want too many other people jumping on these products, as they only work because of cross subsidy. I am not opposed to 1/2-hourly pricing, but the current Agile formula doesn't scale well, especially at lower wholesale prices. Without temporal or regional cross-subsidy and depending on region Agile should never go below something like wholesale, in p/kWh, +6p/kWh overnight, +8p-10p/kWh in the shoulders and+20p-35p/ during the red periods. Of course some of the temporal cross subsidy will be internal to a single customer, but a fair amount of it is between customers. All of the regional cross-subsidies are between customers.

I think the perceived uncertainty in Agile prices will always relegate it to have relatively few customers.

In a way Agile pricing is very much like radioactivity (I was a physicist in a former life). It's very unpredictable and random in the short term but the long term averages are so uniform that we use it to date the age of objects going back thousands of years (radiocarbon dating) or using other isotope ratios to go back millions of years.

Once you have seen the patterns and the way the averages turn out it's a no brainer but most people need Fixed Price contracts or even the SVT as a comfort blanket. I've tried to 'sell' it to other members of my family and friends but so far only my son has gone as far as getting an Octopus tracker and won't use Agile. Ho hum!

 

New question:

Octopus are say that carbon intensity is increasing going into tomorrow

Whereas the National Grid forcast looks like this. Who is right? Or both are right but there are conditions attached to each graph. I.e. the Octopus graph (208g CO2/kWh at 1am) is for me and the NG graph (~62g CO2/kWh 1am) is for the UK.

 

Are you sure you don't need to refresh your page there on the Octopus site? Mine is saying the opposite, and they sent out an email earlier today about there being free electricity tomorrow afternoon for an hour.

 

We must live in different countries. Because of the high wind it should be saying what yours is and that is why I've never trusted Octopus to tell me what the carbon intensity is. I've now raised the issue with Octopus to see what they have to say. It won't be the first mistake I've spotted with their data. The first was three days within joining them.

 

Looks the same as mine. Presumably this is regional, so I'll get 50p for plugging the car in but it won't charge

 

For anyone using IOG and interested in the comparison vs Agile I have this morning checked my last bill period August 8th - Sept 1st and taken into account the free electric I got via the Savings Sessions during that period to work out my average unit cost.

3 Savings sessions during that period netted me £1.94 which brought my averaged unit kWh cost down from 6.97 to 6.60

So not a great amount but IOG users can still benefit from the cheaper Agile periods when a saving session is set up and it helps nudge the rate downwards.

FYI - 8th July to 7th August I saved just 0.87p which gave me a 6.78p average unit cost.

This upcomming bill period so far I have saved £2.62 with credits to my bill for free electric so it will be a tad better.

Now I fully understand that I have house batteries to minimise the day import rate but it does show that for minimal management with IOG you can still get pretty good average unit rates. Also I do import more at cheap rate to offload at the 15p kWh rate.

With Agile you might be able to beat that?

Could those playing with Agile post their averages (if you have time) as it will be interesting to see how your average unit rate fairs over the next few months...

This will help inform anyone reading the thread on what might be possible with the different tariffs.

 

Could those playing with Agile post their averages (if you have time) as it will be interesting to see how your average unit rate fairs over the next few months...

My best is 13.6p and my worst (Jan) was 24.4p over the last 12 months. But I don't have an EV to charge as that would likely bring the average down.
We also never cook between 4-7pm.

 

I do a running YTD figure but I could do a recent moving average if you want to define the time period for that.

YTD values:
A2AHP, All hours except the peak, 4-7pm, 1167kWh (4201kWh heat), 18.4p
Domestic, 24hrs, 2954kWh, 20.0p (includes peak time use)
Car, when it's cheapest generally, 2980kWh, 13.6p
Average (Total cost / Total usage), £1211, 7101kWh, 17.1p

As a comparison with OVO Charge Anytime I'm £46 better off / no boost time, £74 better off 5% boost time, £101 better off 10% boost time. I'm generally better off than OVO CA because I get 'discount' rates for my A2AHP which will not qualify for their ASHP rate.

That doesn't include the standing charge for which the difference is likely to be quite small between energy suppliers compared to the unit costs.

 

My unit cost for the last billing period on the Octopus bill was 12.88 for (I guess that's excluding standing charge) for 842KWH used.

This would have beaten the best I could do on Octopus Go easily and IOG as well - it was a pretty good month for Agile with several plunge pricing and very low periods.

I am mainly only load shifting the EV and PHEV. Dishwasher etc is load shifted in daytime where possible but doesn't have timers.

According to Octopus Compare there were 5 days out of the last 30 where load shifted (optimised) IOG would have beat Agile for me, not taking into account any additional IOG periods that would have been awarded.

 

New question:

Octopus are say that carbon intensity is increasing going into tomorrow

View attachment 215574


Whereas the National Grid forcast looks like this. Who is right? Or both are right but there are conditions attached to each graph. I.e. the Octopus graph (208g CO2/kWh at 1am) is for me and the NG graph (~62g CO2/kWh 1am) is for the UK.

View attachment 215575

If you look at carbonintensity.com or the NESO app you can check both regional and national intensity forecasts.

We also use the OctopusWatch app which shows the forecast intensity and predicts Agile pricing. It will pick a slot for you based on either best intensity or lowest price. Also has some useful reporting tools. Doesn’t really understand IOG though.

It does look like Octopus are using the regional intensities in their Green Estimator. I am currently seeing 27g/kWh for NW England on both Octopus and Carbonintensity (who use NESO's data). Playing through into the future matches Octopus' numbers

 

My total usage on Agile for EV + household (no solar, AHP, battery) thus far in 2025 is 3100kWh costing £460 w/o standing charge (£633 with SC). That's 14.8p/kWh. For 2024 it was 15.1p/kWh I think. 1423kWh of that 2025 figure was the car so just under half.

Octopus Go/Intelligen is 28p and 7p. Assuming roughly half and half usage of house/car that would work out to an average of 17.5p/kWh.

 

My last two IO bills were at an average of 13.4 and 13.8p a unit, as printed on the bill. I have solar but no battery apart from the one in the car.. I do and did timeshift but with low mileage on the car not as much as some people are able to.

Next bill(s) will be on Agile so I'll update when I receive it in 3/4 weeks. No figures yet but have been fairly successfully avoiding heavy 16.00 to 19.00 usage, and should be getting a significant improvement. I've done a bit of V2L - induction hob, air fryer, microwave and the oh so important kettle, but not for real money saving but because I can!

I use very little gas, none at all in the last 6 months, and am considering an A2A heat pump, when the calculations will start all over again!

 

Here we go again.

 

Can I have your rates? Mine are higher here in the SW. Maybe they will fall when Hinckley C comes on line 🤔
But probably not.

 

Rates are in the image.
Say my projected bill will be £2201 but using a home battery reduces that down to around £1100. Will be a bit more now I guess going from 1 EV to 3 on the drive from 2 weeks ago.

 

What is clear, is everyone has unique circumstances and energy usage. Batteries, solar, EV, heat pumps, etc...

Has been useful to see..

 

Here are mine running 2 EVs
We have no battery or solar at moment.

 

Here we go again. View attachment 215660

Here are mine running 2 EVs
We have no battery or solar at moment. View attachment 215693

Good to see that the increases are basically straight pass-throughs of the two external costs (maybe even a bit smaller). The Warmer Homes Discount obligation what changes the standing charge, and the BSUoS cost increase the peak rate. Looking at this, and given the off-peak is still the same, you will come out ahead on that one, just because only a portion of your usage is covered by that increase.

Edit: fix a typo