Monthly usage in payments wildly incorrect

It has recently came to light that the direct debit payment setting cannot be lowered less than 90% of bulbs recommended payment. Whilst I believe that in itself is rather draconian I will ignore that point for this thread.

However, this does leave me in a predicament as this recommended payment is based on some very flawed calculation of my monthly usage.

In the payments section of the app it shows the following:

Monthly payments £95 (info warning)
Monthly usage £156.92
Looking at your current balance and energy usage, we recommend changing your payments to £162.

Clicking on payment settings shows “We recommend £162.35 a month, based on your usage.”
Also “Minimum amount £146.12. This is to stop your account from building up a lot of debit.”

Well I am flabbergasted at this, recommended £1948.20, minimum of £1753.44 annually.

Checking the tariff information on account area of website shows estimated annual cost at a lower £1601.96 but even this is way in excess!

I have been a bulb customer for nearly 3 years with meter readings supplied on a mostly weekly basis so there is no excuse for the above to be so wildly in excess.

My last 12 months actual usage (last 12 statements) totals £1297.81 which is actual £108.15 average monthly usage.

My last statement shows electric projection of £752 and gas of £475, total of £1227 which reflects the recent price changes, thus a monthly projection of £102.25.

Why does the payment system think I use £60 more a month, £720 annually than actual usage and even further disparity from the monthly statement - presumably a separate flawed engine is in use?

Now the only thing I agree with is the difference of usage to recommended as my balance is currently a little low, so £5.43 extra recommended on top of the usage to bring into line in 12 months.

Just to make things worse than it already is, you need to also factor in the approx £230 of FIT payments that feed directly into my bulb account that’s £19.16 a month (should actually be more due to yearly rise linked to rpi each April)

So total recommended payment should in fact be something close to £102.25 + £5.43 - £19.16 giving a total of £88.52 per month.

In summary my £95 is thus already more than ample to keep things in order.

Please look into your flawed monthly usage predictor with urgency!


Current credit £605
Yearly use projection £920
Current payment £25pm
Recommended payment £108pm
Minimum payment £97pm

It doesn’t make sense. It would almost make sense if the credit wasn’t being considered, though even then I’d be paying £244 to £376 too much over the year.

1 Like

Hi @jane37

You’re right, seems like our suggested payments for your account could be a tad off.

I’m going to email you so we can work this one out together, so keep an eye out for that.

Hi @CJ_at_Bulb, any chance of investigating my points too given as yet I have had no response?

Firstly having what looks like three different projections depending on where you look and then choosing to use the worst one for the direct debit restrictions is wrong.

Secondly it not taking into account the FIT payment into the account.


Hi @Phillip_PAL

I’ve had a chance to look over your earlier post, and you’re right - seems we may need to adjust the payments on your account, too.

I’m going to send you an email as well, so if you could check there that would be perfect.

1 Like

So, a manual adjustment was made such that on the tarrif page it shows a more reasonable estimated annual cost of £1204.10 which is £100.34 a month and a temporary block on review of payments such that it does not force the minimum payment.

However the payments page still insists my monthly usage is £127.92 that is only slightly improved so still wildly out (annually £1535.04). Recommended £131 (annually £1572).

Staggeringly the minimum payment it will accept is £117.96 (annually £1415.52) when by my calculations using information from the statement I should only be paying due to FIT contributions £81.75 (annually £981).

My statement for May arrived, with the email showing this useful information (similar, but not identical to the payments suggestions).

|Usage since 25th April|£45.89|
|Average monthly usage|£129|
|Current monthly payments|£95|
|Current balance|£15.98 in debit|
|Ideal balance|£22 in credit|

So the shortfall of £37.98 to make the ideal balance at this time of year explains the recommended payment being £3 above the supposed monthly usage.

In the actual statement it gives a personal projection for electricity of £697 and gas £476, total £1173 which is a projected monthly usage of £97.75.

Why is the payments system calculating a monthly usage 31% in excess of the statement projection?

Just to check my last 12 months actual usage totals £1237.04 (monthly £103.08) this makes sense as the projection is using the updated pricing which are cheaper for my main night usage and gas usage.

Additionally, why does the payments system not take into account the FIT payments approx £230 annually already contributing to my payments?

So we need to recalculate what should be going on.

The payments page should be something like follows:

My monthly usage should be as per statement projection £97.75, giving annual usage of £1173.

Existing shortfall of £38 still needs to be taken into account.

Dont forget the FIT payments totalling £230.

Total annual payments thus need to be £1173 + £38 - £230 = £981.

Suggested monthly payment should thus be £81.75.

So the suggested payment by bulb is actually £49.25 above my calculations. That’s a whopping 60% overpayment!

I would thus like to reduce my monthly payment to £85 however bulbs payment page wont accept below £118.

@CJ_at_Bulb can you take another look and please allow me to reduce my monthly payments to £85.

@Phillip_PAL Thank you for your comprehensive question.

The reason that we don’t take FIT payments in to account is because of the large variation in how much each array will generate. It wouldn’t be possible for us to estimate, given unpredictable variables such as the weather. However if your suggested payment amount is particularly inaccurate due to the FIT payments, we can take this into account on a case by case basis.

I’ll discuss your account with Curtis and get back to you via email regarding lowering the payments further.

Hi @Sophiem_at_Bulb, the fluctuation of FIT payments would be similar if not more predictable than the effects of weather on heating costs?

However as noted the supposed monthly usage is already wrong, the lack of consideration of the FIT payments just doubles the problem.

Hi @Phillip_PAL

I’m impressed by the breakdown you’ve posted above, I’ve enjoyed having a read through the calculations.

I think factoring in the FiT payments there is certainly a strong argument to lower your payments to £85/month. I’m going to send an email separately going into a bit more detail on that.

I understand FiT payments may seem predictable, but they are inherently sporadic on a majority of accounts. Rather than factoring them into our payment calculations, which would be a rather complex process, we keep FiT payments separate. This is more for convenience and transparency, as we can always refund FiT payments directly to our members anyway. However, if your preference is to factor them into your monthly payments to us, that’s also a reasonable way of going about it.

1 Like

I dont normally delve into the figures to such detail, I would normally just check against my position, amount paid and the overall bill compared to the equivalent 12 months ago.

It turns out the estimated usage on the payments page is based upon the prior few years rather than just the last 12 months.

My usage pattern has altered over the course of previous years which explains why the estimate is above my recent usage.

3rd year ago we had two full electric cars both of which necessitated being fully charged overnight, one of which would also require extra topping up during the day on occasion particularly in winter.

2nd year ago the 22Kwh Zoe was changed for a vauxhall Ampera (PHEV) which only had about 11Kwh battery so required topping up most weekdays even in summer.

Last year, the Ampera went for a HEV (non plug in hybrid) so a drop in electric use at the expense of vastly increased petrol costs.

However at similar time I had a small battery storage system fitted which decreased my general daytime usage and allowed in winter for usage of the night time rate for part of the day.

Furthermore my main car charger was also replaced with a zappi, a smart charger which allowed the car to charging to match the rate of solar production, so at weekends, or when working from home particularly over summer I could partially recharge my leaf from the solar.

Very recently the HEV has been exchanged for another PHEV so we can cut down on fuel costs again for short local journeys, however as I am now working from home fulltime this means there is almost always one of the cars remaining at home to make optimum use of the self produced solar.

In past 12 months I have observed my self consumption of solar has increased from 77% to above 90% and the eco7 split has increased from around 70% to over 90% night time usage.

What with the lockdown and Working from home our travel has reduced considerably so the past month of May as an example due to reduced usage and increased solar etc. I have temporary effect of 96% self-consumption providing 82% self sufficiency (last year 78% with 39% self sufficiency)

I will however be WFH longer term, so usage should remain below last year even with the introduction of the PHEV).

My solar panels are approaching three years old and being fairly late on the FIT rates are much lower than the bulk of early installations so perhaps with the lower FIT rate the contribution is perhaps less wildly affected than older installations by the weather.

Hi @Phillip_PAL

It’s interesting to see how your usage breakdown has changed over the past few years; it’s a consistently impressive level of self-sufficiency throughout.

How have you found using the small battery system to be? It seems like that sort of setup is the key to unlocking the potential of renewables - being able to benefit from night rate tariffs in the day certainly sounds beneficial.

The zappi charger also sounds like a really cool idea, I can’t say I’ve encountered it before. Matching the charging of an EV to solar output through the day is a fairly ingenious idea, though, and certainly one I hope becomes more popular with time.

I’ll be interested to see how your consumption patterns are going to change with an extended period WFH. An 82% rate of self-sufficiency is one of the highest I’ve personally come across, here’s hoping that continues.

You really seem to know your stuff when it comes to solar generation, do you have any suggestions for how we could do things better when it comes to our members with solar panels?

The 82% self consumption over the entire month really is just a due to the circumstances combined with an extremely sunny month, but I have had higher such as 30th May which was a full on Sunny day with 98% consumption and 98% self consumption. I will return to this as an example shortly.

Firstly to start with here is May 2018 where the production was just short of May this year.

Back in 2018 you can see the unused green 33% of the production approx 200Kwh was exported unused. I purchased either evening or overnight 720Kwh, which could be reduced by 200Kwh if used more directly. Most of the self consumption was background house usage and a diverter to heat the hot water reducing the gas usage nearly to zero at summer, via the immersion element. I would try and run the dishwasher during a sunny morning and washing machine during the day. Cooking in the evening would consume more than the solar panels produced at that time. If the car was put on charge it would consume approx 7Kw at 32amp where as my solar maxes out in peak of a summer day about 3Kw so the other 4Kw would be consumed from the grid.

The electric cars back then where always charged overnight on economy 7 rate. The solar was mainly to cut down on the day rate consumption which is slightly more expensive than standard (approx double the night rate).

During 2018 I discovered that I could use the granny charger which is a 3 pin plug so charges the car at a safe limit of 10amp. I had an external socket connected which was connected via a relay such that once the hot water was full and there was excess it would turn on the granny car charger, but this although an improvement was hit and miss on changeable days.

Since then in 2019 I had a smart car charger fitted, this can be set to notify the car the available surplus from the solar to track, this had a higher priority than the immersion divertor so prioritises the more expensive car charging over the cheaper gas heating.

Late 2019 I had a small luxpower batter inverter with a 2.4Kwh pylontech battery fitted. My interest in this was to see if it was possible to prepare for the new smart tariffs should a smart meter be fitted, to store energy during the day for use to avoid the peak 4pm to 7pm slot used for cooking.

I discovered that this single battery however only had a peak output of 1.2Kw so although it removed general background usage such as fridge and TV etc. It made little effect on the oven which consumes 2.4 Kw.

A few months later I expanded with a larger us3000 pylontech module with approx 3.5Kwh capacity. With these two batteries the luxpower inverter can provide a peak of around 2.9Kw so can cover the oven or bulk of a kettle.

Interestingly I found that the battery although effectively a 5.9Kwh storage capacity can be partially used multiple times a day. I have in effect discharged over 10Kwh from this setup in a single day. An example shown shortly.

Other changes over the years are:

  • Gradual conversion of nearly all light bulbs to LED (I have about 3 old CFL left or flourescent tube for instance in garage and loft) even the halogen outside light has been replaced once it failed.

  • induction hob fitted in kitchen replacing gas hob.

  • TV upgraded to a larger LED model which gives off less heat and less power hungry than the prior smaller backlit LCD model.

  • When washing machine tumble broke last year, replaced for one with a delay timer.

  • when the dishwasher broke the other month, replaced for one with a delay timer.

The above means the battery needs to output less during the evening so lasts longer, the dishwasher and washing machine are set for overnight usage for most of the year.

When cooking in the evening I try to start with boiling kettle first when required and waiting for it to finish, only then turn oven on, thus lowering and spreading my load such that solar and battery have better chance of providing the demand.

Notably the induction hob at partial power the battery struggles to follow as it is a rapid fluctuating on/off demand every few seconds, whereas the oven fluctuates every few minutes.

I am happy this setup would cover the peak 4pm to 7pm for my usage with a little careful management. An extra battery module would improve its peak output to 3.5Kw and also increase its storage, but each increment does get less beneficial cost wise depending on your circumstances.

At this stage battery storage is not really much of a cost saving exercise - particularly for the short term, but it does help spread the load greatly for the grid and makes me feel better, and open up the possibility of moving to a more complex tariff once smart meters are available.

Now back to my example of 30th May 2020 which achieved 98% self consumption. Previous day was a very sunny day and the battery was nearly fully charged as I was charging car until just before 4pm.

I knew it was going to be an extremely sunny day and we had run out of washing, so nothing was on overnight and the car was nearly full so it was left on eco+ setting on the zappi to charge only when the solar was getting strong.

Afterwards we took the PHEV car out and returned, I put this car on fast charge it charges a max of 3.2Kw (16 amp)

The batteries which had been storing power during our absence provided the difference of what the solar was producing.

We went out shopping later and returned popping ut back on charge, then a short trip even later in afternoon and recharged. The car was effectively full before we started cooking for tea in early evening, this was mostly the HOB and you can see there was some limited purchase and export to the grid due to the fluctuating hob.

It’s only on these multiple occuring sunny days where for my own circumstances the battery is also powering me overnight. I aim to only import grid use overnight, yet ensure the PHEV is topped up to limit or avoid consumption of fuel on short journeys.

Nearly 18Kwh went via the zappi to the cars battery which would have provided about 65 miles purely by solar

Here is the battery monitoring showing its charge and discharge. You can clearly see the continuing fluctuation caused by the fridge.

The State of charge of the battery

Finally the Month of May 2020 to compare against the original 2018

Here you can see some days where the green blue and red are extremely close showing barely any import or export with plenty of consumption.

When the red is lots higher this was consumed overnight.

Today my leaf was recharged but as it has a larger battery this was mostly from travels undertaken a few days ago, we have had so many days of sunshine and rarely been out today that I have been recharging my leaf.

It must be noted that this has been an exceptionally sunny May with May June and July being the peak sunny months.

In December and January the solar production is very low such that 4 sunny days in summer is equivalent of that produced over January and December combined!


Hi @Phillip_PAL

All I can say to the extent of the data you’ve taken the time to share with our Community is wow! I’m not sure we’ll be able to give the response that your contributions to this thread merit, but reading through the data was really interesting and I think you broke down your usage habits in a really helpful and accessible way.

Lots of what you’re doing, for example your vastly reduced demand on the grid by only importing from it overnight, really aligns with the direction Bulb is looking to take in terms of becoming an energy manager. Some of the ways we’re doing this are our EV charger trial, improving and increasing the smart data we give members access to in their Bulb Accounts, our partnership with Samsung Smartthings, and our newly launched Carbon Calculator, which I’d love you to try here

One things that really comes through is shift in relationship between the home and energy generation, into something more symbiotic and that really gives the data meaning, at least to me. Every change you’ve made, from usage patterns, to replacing appliances has been an upgrade, whether driven primarily behind impact reduction, or long term costs saving this is really admirable.

You mentioned using the battery to partial capacity several times a day, to effectively exceed the 10kwh storage capacity across the day, just in a staggered manner is a really cool workaround. One thing I’d be interested to know, is how much conscious effort do you think it currently takes you to think of these solutions, research smarter appliances, or charging systems and generally manage your home in a better way? I think the next step may be to make this simpler to do for the wider population and feedback on current tech from people like yourself will prove really insightful.

I’d also be interested to know where you accessed / produced the graphs you’ve shared with us and if you’ve developed an API for any of the above. The first graph you included when going back to talking about 30/05/20 is particularly visually satisfying. I also think the difference in solar production between seasons, while probably not something that should surprise me, is astoundingly stark. I imagine when I do buy a place and have solar installed I’ll obsess with weather forecasts.

Again thanks for sharing, there’s not much Bulb loves more than energy data.

1 Like

I have seen others who have been able to drive complete dashboards and overall control / improved automation via bespoke tools via the api of other products.

This is beyond me at present but does interest me for the future if I get time.

Here is my monthly generation over nearly three years. I have a split system 11 panels roughly east and 5 on the west side. Half of the east side is heavily shaded by next door roof until mid morning - particularly so in Winter, the west side works more afternoon into early evening but is then shaded by trees.

If Ihad known better i could relocate 1 or two panels on the east side this would improve winter generation slightly.

The system required optimizers behind each panel to cope with the shading.

I paid extra to have the solaredge consumption monitoring system fitted, this plots the average in 15 minute chunks but i only use this for historical nowadays.

The luxpowertek battery inverter shows the most frequent history of grid, solar, battery and thus overall consumption. The graph portal is free via luxpowertek. It gives this information every 2 minutes so can clearly identify the fridge cycling.

The battery can be reprogrammed via web interface (but klunky) to set it to charge overnight I do this roughly September to March.

The zappi smart car charger from myenergi is what I have invested heavily into adding a hub to enable remote control and monitoring via web access also a harvi with ct clamps to monitor the battery and the immersion diverter. This allows a more instant overview every few seconds of the entire system, its app is improving greatly but as yet provides no history for the battery. I use this to set timers or change modes on the zappi between fast, eco (charges car at a minimum of 1.4Kw unless more surplus available) and eco+ (this suspends car charger completely if cloudy conditions arise or the oven us turned on for instance to avoid depleting the battery).

Forgot final picture from myenergi my last weeks consumption, this shows I do occasionally import from the grid overnight even in these sunny conditions!

Hi @Phillip_PAL,

This is really impressive stuff :electric_plug:

Your Zappi smart car charger sounds like the ideal setup.

Having just gone through what you wrote, what really stood out to me was the way your solar panels and batteries have cause a slight behavioural change when you cook in the evening. Behavioural changes towards energy consumption is something our team have spoken a lot about recently. Over the last few months, I’ve some really interesting and qwerky ways people are saving energy through new habits.

That being said, I haven’t heard much in terms of lowering and spreading a load so solar panels and batters have a better chance of providing the demand.

Have you noticed any other behavioural changes that you have made often the last few years in order to spread the load on your solar panels and batteries? :battery:

When I first heard of smart tariffs and the peak 4pm to 7pm period, I did a run through of my usage particularly in winter and discovered it would be more costly than the current eco7 setup, unless the peak was avoided and not cooking Outside this window is not an option. Hence the looking into the battery solution which has worked really well much better than anticipated for a small system.

Before I had timers on the washing, I would try and start the wash around 11pm and then the tumble cycle would start at around 00:30 for the night tariff for the winter period. The dishwasher I would start early in the morning before shower when I let the dogs out at 05:30.

The main hit is the kettle at 3.3Kw. So i put the toaster on after the kettle.

The bigger the battery system though does allow greater flexibility to not need to follow these as religiously but to have a general awareness helps.

The kettle is definitely the big one.

I’ve heard so many energy saving tips that revolve around our infamous kettles. A few people have told me they store the extra hot water in a thermost either to make another cup of tea later or to use for the washing up. Pretty clever stuff :tea:

In a similar vein, we shared on Instagram that by only boiling enough water for each cup of tea you can save around £6 a year. I’ve been trying to do this one myself, but took a bit of trial and error to match the water level with my favourite mug.

1 Like