Overestimation by shell energy

ClareAshton

We have been with Shell Energy since February 2022.  That's less than a year, so why are they saying our prediction from November 2021 to November 2022 is 5000kW?  We haven't been with them for a year

Apparently, they base your energy bills on the last 12 month period.  If you haven't been with them that long they look at your postcode and estimate your usage on an average in the area.  There are two of us, my husband and I.  We have smart lighting, Hive heating and an electric car.  We do our washing at night and we charge the car off-peak.  We rarely use the TV.  All we use is a laptop and two PC's.  So how can they compare our usage to a family of four in the same size house and say we have the same energy demand?

[Deleted User]

ClareAshton said:

We have been with Shell Energy since February 2022.  That's less than a year, so why are they saying our prediction from November 2021 to November 2022 is 5000kW?  We haven't been with them for a year

Apparently, they base your energy bills on the last 12 month period.  If you haven't been with them that long they look at your postcode and estimate your usage on an average in the area.  There are two of us, my husband and I.  We have smart lighting, Hive heating and an electric car.  We do our washing at night and we charge the car off-peak.  We rarely use the TV.  All we use is a laptop and two PC's.  So how can they compare our usage to a family of four in the same size house and say we have the same energy demand?

They are forecasting ahead. Their data collector uses past consumption which the industry has held, in respect of some properties, for over 70 years.

Estimated usage is used to set a monthly DD payment. The monthly payment is added to your energy account as a credit and charges for energy used are deducted from the credit balance. Remember, electricity usage in Winter is higher than mid summer.

PS 5000kWh a year is not high if you have an EV: peak or offpeak, the EV is agnostic.

[Deleted User]

Give them accurate readings and you will get accurate bills.  It's just as simple as that.

You might end up with a credit in your account if they are indeed overestimating, which could be useful if/when prices go up again.

If the readings are accurate, you won't be overcharged.

MattMattMattUK

ClareAshton said:

We have been with Shell Energy since February 2022.  That's less than a year, so why are they saying our prediction from November 2021 to November 2022 is 5000kW?  We haven't been with them for a year

Why do you think that is an overestimation, on what are you basing your opinion?

ClareAshton said:

Apparently, they base your energy bills on the last 12 month period.  If you haven't been with them that long they look at your postcode and estimate your usage on an average in the area.  There are two of us, my husband and I.  We have smart lighting, Hive heating and an electric car.  We do our washing at night and we charge the car off-peak.  We rarely use the TV.  All we use is a laptop and two PC's.  

They do, but they have vast amounts of historical data that means even with less than twelve months they can fairly accurately estimate future usage. You also do not sound like low users as you have an EV.

ClareAshton said:

So how can they compare our usage to a family of four in the same size house and say we have the same energy demand?

They are not comparing you to a family of four, they are basing your usage on data since February as well as historical data which correlates against the months that they do have data for, regional usage, home size etc.

ClareAshton

The car needs charging twice a week if that - it is small battery - 24kWh

The 12 month quote is useless we have been with them 10 months, you can't give you annual usage with 10 months data.

Our PC's are on maximum 2 hours a day

MattMattMattUK

ClareAshton said:

The car needs charging twice a week if that - it is small battery - 24kWh

Two charges of a 24kWh battery works out at approximately £17.50 per week, so an average of £75.80 on your monthly bill. That is based on full charge and 100% charge efficiency, it is likely that the charges are not full, but also that the efficiency is closer to 85-90% depending on the exact setup. 

ClareAshton said:

The 12 month quote is useless we have been with them 10 months, you can't give you annual usage with 10 months data.

Actually you can, ten months is almost certainly enough to extrapolate twelve months usage, especially with the vast amounts of data over many years that the industry has.

ClareAshton said:

Our PC's are on maximum 2 hours a day

One thing in isolation means little, which is why people go into granular detail. The fridge-freezer can use 2-8 kWh per day depending on model, usage, even location in the property. Washing at 40c or 60c uses a lot more water than washing at 20c, some people only wash full loads, others wash partial loads several times a week, some people run the dishwasher every day, others only once or twice a week etc. 

With ten months usage the provider will be able to see where you align with others users over the last ten months and use that to get a reasonable estimate of where the remaining two months of the year will be, especially when they bring in additional data such as region, weather patterns, even time of day usage, they all give them the ability to estimate a full twelve months based on the previous ten. 

tim_p

ClareAshton said:

The car needs charging twice a week if that - it is small battery - 24kWh

The 12 month quote is useless we have been with them 10 months, you can't give you annual usage with 10 months data.

Our PC's are on maximum 2 hours a day

Should be relatively simple to get an idea of how accurate that 5000kWh estimate is. What was the meter reading when you opened the account / signed up with Shell and what is it today?

MattMattMattUK

MikeJXE said:

MattMattMattUK said:

ClareAshton said:

We have been with Shell Energy since February 2022.  That's less than a year, so why are they saying our prediction from November 2021 to November 2022 is 5000kW?  We haven't been with them for a year

Why do you think that is an overestimation, on what are you basing your opinion?

ClareAshton said:

Apparently, they base your energy bills on the last 12 month period.  If you haven't been with them that long they look at your postcode and estimate your usage on an average in the area.  There are two of us, my husband and I.  We have smart lighting, Hive heating and an electric car.  We do our washing at night and we charge the car off-peak.  We rarely use the TV.  All we use is a laptop and two PC's.  

They do, but they have vast amounts of historical data that means even with less than twelve months they can fairly accurately estimate future usage. You also do not sound like low users as you have an EV.

ClareAshton said:

So how can they compare our usage to a family of four in the same size house and say we have the same energy demand?

They are not comparing you to a family of four, they are basing your usage on data since February as well as historical data which correlates against the months that they do have data for, regional usage, home size etc.

So how do you work that out. ? 

How do you work what out, estimates of annual usage? You run algorithms/formulas on large data sets, in the hundreds of thousands to millions range, used properly they can spit out remarkably accurate predictions for the majority of people, although there will always be outliers. 

MikeJXE said:

OVO are saying I will be in debt by £740 by next December,
They don't have any history I have only been with them 2 months
I do have history over 2 years and so does my last supplier who estimated my current DD
I also run a spreadsheet and I am using less than forecast

Two months data would generally be too low to gain a valid estimate, that is only 16.6% of the year so prediction accuracy would likely be low, ten months of the year is 83.3% of the year, predictions would likely be accurate.
You can submit regular meter reads and they will then account for that, but as they only have two months data they need to make a prediction based on the data they have, although in those circumstances they are normally much more flexible with customer predictions, especially if you talk to them and explain why you are making your current prediction.

MikeJXE said:

I can prove my estimate of my future use OVO cannot

You cannot prove an estimate, that is a contradiction. You may believe that your estimate is likely to be more accurate, the fact that you have two years of historical data and they have two months is likely that you will be far more accurate provided your workings are correct. 

MikeJXE said:

IMO they are trying to get everyone in credit by the end of the winter and let that credit increase over the summer so they can Keep a good bank balance. 

They are not required to do their best to stop customers getting into debt at any period of the year, they are also required to not allow customers to run up excessive credits, but the regulator errs on the side of credit rather than debt. In theory they are aiming for zero balance at the twelve month account review, as an example EDF refunded a chunk of my credit and lowered my monthly Direct Debit, mileage will vary. 

MikeJXE said:

Please in my case don't tell me I am wrong and feed me a load of ###,  I can prove I am not 

I think you are mistaken, I am not sure what you feel you can prove either, other than 16% of a year does not give an accurate prediction of a whole year. 

MikeJXE

 MikeJXE said:

Please in my case don't tell me I am wrong and feed me a load of ###,  I can prove I am not 

I think you are mistaken, I am not sure what you feel you can prove either, other than 16% of a year does not give an accurate prediction of a whole year. 

I can prove the estimate from SSE who I had been with for 2 years and predicted on their last bill  is way lower than the estimate provided by OVO after only 2 months. 

So if the energy companies have this wealth of data why the discrepancies especially when OVO  have recently took my account over from SSE

Therefore the DD that they propose to force on to me is way too high. 

They might be able to spit out remarkable estimates for the majority of people but not everyone.

I am not average, You can have a 100 different size potatoes in a bag and find an average but that does't mean there are 2 the same size 

[Deleted User]

Proving that one estimate is lower than another is possible.

Proving that either estimate is wrong is not possible until the entire period covered by the estimate is completed.

Showing that one estimate is more likely than the other to be correct is actually what you are talking about.

MattMattMattUK

MikeJXE said:

 MikeJXE said:

Please in my case don't tell me I am wrong and feed me a load of ###,  I can prove I am not 

I think you are mistaken, I am not sure what you feel you can prove either, other than 16% of a year does not give an accurate prediction of a whole year. 

I can prove the estimate from SSE who I had been with for 2 years and predicted on their last bill  is way lower than the estimate provided by OVO after only 2 months. 

Ok...

MikeJXE said:

So if the energy companies have this wealth of data why the discrepancies especially when OVO  have recently took my account over from SSE

As previously pointed out, two months is not long enough to make an accurate prediction, they only have two months of data out of twelve, that will not lead to an accurate estimation. The original complaint was that ten months out of twelve is not enough for an accurate estimate, which is is. You can provide them with the usage from your previous provider and hey will likely agree to adjust the amount accordingly. 

MikeJXE said:

Therefore the DD that they propose to force on to me is way too high. 

You also have the option of whole monthly bill by Direct Debit if you wish, or alternatively you provide them with the figures from your previous provider and see if they are willing to change the amount.

MikeJXE said:

They might be able to spit out remarkable estimates for the majority of people but not everyone.

I am not average, You can have a 100 different size potatoes in a bag and find an average but that does't mean there are 2 the same size 

It does not, but if you have 1.1 million potatoes there will be tens of thousands of the same size, if you use data from use 83% of their growing cycle you can fairly accurately predict yield and quality. The original subject was Shell, they have around 1.1 million customers, they will use the customer profiles that match the OP's energy usage for that ten month period, including things like EV charging, usage patterns etc. to establish a reasonable estimate for the remaining two months of the year. No one is claiming that an annual prediction based on two months would be accurate, there are too many variables and not enough data, but estimating twelve months usage based on ten months is possible with a fairly high degree of accuracy.