Home energy

How I monitor it

For a long time I have been interested in the weather, not just the weather but how it affects the amount of energy that my house uses.

I have lived here on the Isle of Wight since 2004 and have taken gas and electricity meter readings each week along with max & min temperature readings each day.

If one is plotting coldness we need a unit of measurement that is bigger as it gets colder. The unit is known as a Degree Day or HDD (Heating Degree Day) there is also a Cooling Degree Day CDD

It was not until 2021 that I worked out how to best use my data. I brushed up my coding knowledge and made an App that analyses the data. The App, known as DD-Plot reads in a text file, calculates the best fit parameters and plots the points on a PC screen.

Last year (2022) I asked a few friends if they would note their gas meter readings each week. This they did from September 2022 to December 2022


The results are shown on the left. It is not fair to compare the differences between the five properties as all have their own specific differences; type and size of house, size of family, whether gas used for cooking and how the DHW is heated.


The image shows a HDD v m^3 plot for the 10 week period Oct to Dec 2022.



There are five properties







But remember, these results are for the same 10 week period and so have the same HDD.


I am sure that many folk see me as the "Mad Professor" type, with a crackpot idea but I am only following guidelines made some time ago by professional bodies.








Mad professor or not, I am doing practical work. Early in 2022 we could all see the writing on the wall. I changed the timing of the boiler schedule and adjusted the room thermostat and TRV's.


Throughout the remaining months of winter / spring 2022 I took the readings and ran my software. Now this is not a proactive process but a backward facing method over a long term. However I think it works.


The results of the two year comparison are shown below.






The first year's data finishes at point number 14, points 15 & 16 are zeros

 the first point in the second year is number 17.



Yes it is time consuming, here is the method:-


Once a week



    

  1. Read the gas meter
    2. 

  2. And find the HDD value
 from here
for the prior week.
    3. 

  3. Make a note, Away that week, Guests etc. 
    4. 




Using Excel which will show the graph (Not Google Sheets which will not)




Create a spreadsheet with the following columns




Date Meter (m^3) Diff HDD Comment


week 10 1000 50 35 This


week 11 1100 100 70 That


week 12 1200  --  --  





Note the Diff column is the difference between each week's meter reading (week 12 - week 11) the HDD value in this case 70, will be in the week 11 cell. It's the HDD for the week of week 11.


Then make a scatter chart HDD v Diff


My software also calculates the kWh used in the period. You can do this, multiply the total gas consumed by the calorific value ( I use 11.1922 but it can vary). For the cost, multiplying  the kWh value by your tariff figure (approx ₤0.10) will give you the cost of gas. To which you should add the cost of the daily standing charge for each day.


I am pleasantly surprised by how sensitive this method is. I also plot daily readings. Below is the plot for the month of December 2022. Point 25 shows that I used the Gas oven to cook Christmas dinner.
 
I normally use a halogen oven.










Perhaps you may like a copy of the .CSV file that I used.


The format is :-
DD,   m^3,   Comment   
 click to download    (dec-22.csv 1k)    Then open it in Excel





Isle of Wight local information

The Island's climate is milder than the rest of the UK therefore my HDD's are not typical.







I have looked at the last five years of my data and taken the average HDD for each calendar month.


I then calculated the percentage of the year's total HDD for each month.


The image shows that the four months from November to February amount to 64% of a
years HDD but I guess we already knew that.








Method limitations

The concept behind the HDD calculation is based on an the fact that there is an even, almost sine wave shape to the daily temperature curve. But this is not always true. Some winter days start with a sunny but very cold morning and finish the day dull and rainy. The approximate calculation for HDD in one day is the difference between the base temperature (15.5 deg C) and the average temperature of the day.


  You can say that the ave temp is (max + Min) / 2 but if the day is skewed
there will be some inaccuracy.



  I have calculated HDD here over the last 4 months using a data logger which takes measurements every hour and there is very little difference between the approximation and an hourly method.



Below is a graph that shows the comparison between the ten year average and the present heating year (2022/23). You can see that this year has been much the same as average until week 20 then warmer less cold until week 30. Then back to average.














A bit of history


We moved into this house in early 2004,  I now have Degree Day records for all 19 years


But what to do with them, how to display / analyse  them?  Before that I need to set a standard. I will only consider the heating year,
early October to late May  (34 weeks in all).


Each week of the year has HDD, Gas used  and a comment. The comment should mention holidays,  guests etc,




Here is the data for the Years 2004 to 2023









Columns A to D I hope are self explanatory.



Column E is the ratio of Gas : HDD. This is easy to calculate from your gas meter and public HDD data.
However if the household cooks with gas or the DHW is heated by gas and it is a large family  then this ratio will not be meaningful.


The Slope and Intercept are  measures of the line that our heating system's produces, see more below.


Columns H to J all show Degree Day data, it is here that we can understand  how cold the year was;  Mild, Cold or Very Cold




I have decided that a week with more than 66 HDD is a "Cold week".




I need a figure that differentiates between a long cold winter compared with a mild winter. Let's look at the HDD value of the coldest week multiplied by the number of weeks greater than 66 D-days (>66 count).  This is not related to a real thing but it does  give an exaggerated figure for a long cold winter.  This is what I have called the  product.


This figure does seem to point out the cold years, shown in red. And shows that although the year 17-18 has the coldest week (107)
it was the year 12-13 that can be considered a really cold year.



Now to the performance of my heating system

Let's look at the last three years, as it is here that I have had the tools (smart meter) and, with the cost of energy increasing,  the need to control the heating better.


Note that the Ratio has been reduced in value from; 1.26  then 1.08 and last year to 0.99


See also that the years either side of 21-22 are colder and that the amount of gas used in the year 22-23 is lowest of all.


Although a crude measure the ratio,  Gas : HDD is the one to watch and can be achieved with a low tech method.

A graphical Approach

Whether High or low tech,  a picture can paint a thousand words.


The slope of the line will depend on how you and your house use and control your  heating.
Your aim is a low value but not so low that there is little change, mild or cold.

See the example of 5 properties earlier in this page.


The Intercept should be close to zero and is an indication of your DWH heating.


A notepad and some  graph paper is all that is required.


I said that I now use a smart meter. This allows me to read the gas meter daily from indoors. All my data before 2021 were taken once a week
by me going outside, in the dark and cold, torch in hand. So much easier now but otherwise no difference.

Why ?


We hear so many times that folk use their monthly payment as a year to year comparison. Clearly we can not do that now as the price of fuel changes. So we should compare gas use, it is easy and on the bill. But NO each year is different due to the weather.













This page was updated on

12 June 2023 by Terry Rawkins