Jump to content

SteamyTea

Members
 View Profile  See their activity

  • Posts

    23368

  • Joined

  • Days Won

    190

 Content Type 

Blog Entries posted by SteamyTea

  1. SteamyTea
    Been a bit bored as I have not been able to concentrate on anything much recently, but life is getting back to normal.
     
    I read on here a lot about the advantages of insulation, airtightness, MVHR etc, but this misses a few points.
    Form is one of them, a simple cube is a pretty good shape for thermal efficiency, and the bigger it is, the better it looks.
    So I thought I would knock up a very basic spreadsheet that can be used to explore the differences between size, form, thermal properties (just U-Value), air changes an hour for a cuboid.
    This is basically to just show the ratios and theoretical power transfers.
    Cells B3 to B11 can be changed with the snapshot results shown in cells B13 to B17.
    Below that are some data arrays that show ventilation and fabric losses, and ratios for surface area to volume and fabric to ventillation for different form factors for the cuboids.
    Not sure how useful this will be to anyone, but it does put numbers to changes i.e. you can change the ACH and U-Value and see which will have the greater effect.
     
    This is a very limited scope spreadsheet, so a lot of interpretation is needed.
    It does produce some charts though.
     
     
    Compare U-Values and ACH.ods
  2. SteamyTea

    Off Grid Thoughts

    By SteamyTea,


    Off Grid
    Off Grid Challenges
     
    Recently, there has been a few people talking about off grid living. This is an admirable and romantic idea, and something I would like to do myself.
    Then reality kicks in.
    First we must agree what we mean by off grid. To me it means not connected to main services, but usage of public services i.e. roads, domestic rubbish, healthcare, education, policing, food, clothing etc is allowed.
    Basically it comes down to water, waste and power. I have come to these limitations because I cannot live without the others. No one want to see me in handmade clothes and washed only in spring water.
     
    So the first thing to look at is how much energy I actually use, and when I use it.
    Luckily I take a keen interest in this and log my electrical data every few seconds (a mean of 8.5 seconds). Internal temperature data is also logged.
     
    Initially, to keep things simple, I work with monthly data, but will drop down to weekly, daily and half hourly data when needed.
    The chart below is the big picture of what I use.
     

    Chart 1
     
     
     
    I am generally quite happy with my overall usage and internal temperature. This winter I am experimenting a bit with the temperature as my working hours have changed after 15 years of working evenings I have gone back to working days.
    As Chart 1 shows, January to April shows I have my storage heaters on and use a Mean Power of between 0.7 kW and 0.4 kW, then it drops to 0.36 kW in March, then down to 0.25 kW until the end of October.
    As you will have noticed, there are two Mean Power lines. One is the ‘normal’ interpretation of the Mean i.e. the sum of all readings, divided by the number of readings. This obviously includes the minimum readings, which are 0 kW and sustained maximum 4.33 kW. ‘Sustained’ in this instance means for at least half an hour. There are times where the Maximum Power peaks at 13.16 kW when looking at half hour data, but this may be for just a few seconds or minutes.
    By filtering out the 0 kW times, one gets a more realistic idea of what is actually being drawn and when it happens. This is important as it can help when choosing storage and delivery systems.
    Without changing anything, including usage behaviour, I could get an inverter that has a maximum power delivery of 14 kW and know that all my usage would be covered. Storage, for one day, without any inputs, would need to be at least 26 kWh once 20% efficiency losses are taken into account. The 20% losses are based on a ‘best guess’ as this is a very variable percentage based on different power draws, battery charging and discharging temperatures and the state of charge they are at.
     
    So what do I use my energy for. Luckily, being on E7, I can easily see what is used for water heating and what is everyday usage.
    By looking at my April to November night usage, I can get a fairly good estimate of my domestic hot water usage.
     
     

     
    Chart 2
     
    Chart 2 above shows the half hourly usage between April and up to November, when there is no space heating on.
    The mean usage over that period is 4.2 kWh, so apart from the washing machine usage and the fridge switching on, it is fair to assume my daily water energy usage is around 4 kWh/day. This is bourne out my my higher than average water usage. One of my many failing is that I like a bath, every day, and would have two a day if I did not curb my enthusiasm.
     
    If I fitted a heat pump, to heat the water, I could probably reduce that down to 2 kWh/day. Or if I took showers, less than 1, but I dislike having a shower, though they are good at getting the day’s fat, blood and grizzle off my body after work.
     

    Chart 3
     
     
    Chart 3 shows the same data, but for January and February when the storage heaters where on. The daily mean, for space and water heating, has increased to 12.25 kWh/day, so 8 kWh/day are for space heating. This works out as a power delivery, for my house, of 27 W.m-2. Using a heat pump could reduce that by a factor of 3, so less than 2.6 kWh/day or 9 W.m-2.
     
    Looking at the mean internal air temperature, I see they are within 0.5°C. This is good as it shows that my heating regime is working well and does not need adjusting.
     
    So having got my usage figures, and estimated some usage figures if I changed to a heat pump, what can be done about generating energy on site to cover approximately 8 kWh/day.
     
    My house is small, and the roof will only support, at the very most, 3 kWp of photovoltaic. It is also less than ideal facing with the optimal side facing South West.
    Looking at PVGIS to get an estimate of what I could generate, highlighting December because it is the worse month and with similar usage to January, it shows that there would be a total generation of 24 kWh.
     
     
     
    That works out as around 0.75 kWh.day-1
     
    It is not until April that I could cover my usage, and by October a deficit would be showing.
     
    The deficits are in the table below.
     
    Month
    Usage /kWh
    PV Generation /kWh
    Deficit
    January
    248
    31
    -217
    February
    232
    58
    -174
    March
    248
    129
    -119
    October
    155
    83
    -72
    November
    180
    38
    -142
    December
    248
    24
    -224
     
    Whiles the above energy deficits are not that large, they need to be covered.
    Even if a battery storage system was installed, without the generation capacity, regardless of how spasmodic the generation, it would still not be covered.
    The only realistic generation method is to use a small generator.
    Using December’s data, as it is the worst month overall, on average, a 2 kW generator would have to run for 4 hours a day once efficiency losses where taken into account. During these 4 hours, a battery system of 26 kWh capacity, could be efficiently charged with 8 kWh of energy. By having an oversize battery storage system, more effective charging and discharging can take place, and the system will have a longer lifespan. It also allows for some days to probably not run a generator at all depending on the weather.
     
    My choice for a generator would be bottled gas (LPG). While diesel may offer a small improvement in efficiency, they are noisy and if the stored fuel gets some water in it, can be expensive to rectify or repair. Gasolene to LPG is a tried and tested conversion.
     
    Ideally a combined heat and power (CHP) unit would be used as these offer the best possible efficiency with about 30% of the fuel input turned into usable thermal energy and 20% into electricity. Unfortunately there are no easily available small CHP units or around 5 kW total output.
    This would mean that a DIY solution would have to be made. This would be an interesting project. There are some small capacity, water cooled, twin cylinder motorcycle engines that may lend themselves well to this application. There are also cheap, permanent magnet, low speed, direct current current motors that can be driven as a generator.
    Noise would be the biggest problem with a generator, but as it may only run for a few hours a day, then it can be used during the daytime. It can also be buried in an earthen bank, with secondary inlet and exhaust systems fitted. Modern cars are very quiet at low revolutions, no reasons that a modern motorcycle engine should be any different. My car, and old diesel is quite quiet at tickover, and it is using 1 kWh of fuel per hour.
     
    So to conclude, while it is not possible for me to be ‘off grid’, with a larger, more isolated property, and the use of a generator, and about £25,000 of investment, I could be off grid.
     
  3. SteamyTea

    My Enery Monitoring

    By SteamyTea,


    My Energy Usage January 2019 to Octobe 2024
    As many of you know, I like to measure everything.  My energy usage is just about the top of the list, and as I have got a new PC with a bit more power than the old one, I have been able to analysis several years data in one go.
    The main areas I have been monitoring are electrical energy usage, in my all electric house, internal and external temperature, and a weird one, zero power draw times.
    For the last decade or more, I have been using a CurrentCost Envi, this has an optical sensors that counts the number of 1Wh pulses off the main electrical meter.  It is what gave me the idea to design a similar one.
     
    Over the last couple of decades, I have slowly been improving both the fabric of my house by replaced old glazing, but not frames, added 'lockout times to limit the E7 window, bought lower powered electronics, added insulation and improved airtightness.  The biggest thing I have tried to do is to make sure that there are no parasitic loads, this is where the Zero Power Draw metric comes in.  This got me thinking that if I ever went for battery storage, which usually has a minimum power draw to start, with 200W often being quoted, would it be of any use to me.
     
    In August 2023, I foolishly agreed to a smart meter being fitted.  Now I have no objection to a smart meter, in fact, I quite like them.  The problem was that my energy supplier EDF, fitted a dud, then spent months sending me incorrect bills, took weeks to replace it, changed my contract without informing me, and eventually had to give me a couple of hundred quid as compensation, so not all bad.
    What this has allowed me to do is to calibrate my CurrentCost meter.  Since the beginning of the year, which is the only time I have reliable smart meter readings from, until the beginning of this month, October 2024, my smart meter has registered 2383 kWh, while my CurrentCost has registered 2349 kWh.  Taking one from the other, dividing by the smart meter amount, multiplying by 100 gives me a percentage difference, 1.4%.  So not exact, but can probably be accounted for for the few times that I have disconnected the CurrentCost (moved it, sender battery went flat, other disconnects).  One year, I shall have a trouble free, 100% full readings.
     
    So now some charts.
    This first one is a time series since the beginning of 2019.  There is a little break in the outside air temperature readings (was using a DHT22 and it probably got wet in a storm).  There are times when I get seemingly high readings which are caused by the sensor being hit by full sunlight, but for 22 out of 24 hours a day, the readings are good, and I can filter out the over temperatures if I need to.  I have a plan to improve this by making a shield, but going for a coffee is more fun than doing 'engineering'.
    When we had the energy price shock back in 2022, I added triple glazing to my windows.  This was a cheap option (about £120) and after the first winter I noticed that I did not have to turn my storage heaters on until the mean outside air temperature went below 9°C for a few days.  This was a degree lower than before.  I also noticed a drastic reduction in sound transmission from outside.  There is an energy reduction that is quite clear on the charts, even though there is missing temperature data, though luckily during the summer.
     

     
    This second chart shows the mean power by time of use and the percentage of time that the house draws no energy from the grid.
     

     
     
     
  4. SteamyTea

    Adding extra mass to buildings

    By SteamyTea,


    Thermal Testing
    As many of you know, I have been sceptical that adding extra mass to a building will stabilise the temperature.  This basically comes from when I studied this back in 2008.
    Just to reconfirm my suspicions that regardless of construction type, the air in a building will react more to external inputs than internal inputs i.e. solar gain, ventilation and a heating system versus thick stone, brick or concrete walls.
    To retest this idea, I worked out the amount of energy that is needed to change the air in one of my rooms, then calculated how much water is needed to match it.
    Then I started measuring  temperatures over the last few days.
     
    Now I can bore you all with very detailed statistics, but it boils down to what I showed 16 years ago, it basically makes no difference.
     
    The headline figures are that the mean air temperature was 16.7°C with a range of 6.2°C for the air in the room, 6 litres of water, which needs the exact same amount of energy to change by 1°C had a mean temperature of 16°C and a range of 25°C and double that amount, 12 litres, had a mean temperature of 16.3°C and a range of 2.9°C.
    The bigger range, which some will interpreted as instability is caused by natural air changes and heating input (I have storage heating, so does not modulate like a properly set up combustion or heat pump system).
    When looking at the more stable stable times between 10 AM and 3 PM (when I am usually out) the mean temperatures are, for the room 17°C (range 0.1°C), 6 litres 16.1°C (range 0°C), 12 litres of water 16.4°C (range 0°C).  So allowing for instrument accuracy, about the same range, but the masses are actually cooling the air.
    When I looked at the rate if change in an hour, all three were the same at -0.1°C/hour when cooling, and when warming, the air reacted faster at 0.3°C/hour but only over the time the energy is being inputted (two hours in my case). The water masses are equal at 0.1°C/hour.  So the air responded a bit faster overall, but not much over 24 hours.
     
    All the above confirms what I researched 16 years, adding mass to a building will make the mean temperature lower.
    If you want to stop overheating, change the window design, no need to fill the walls with concrete, brick or block.  That is barking up the wrong tree.
  5. SteamyTea

    Heat Capacity; What it is.

    By SteamyTea,


    Thermal Testing
    Heat capacity is a really simple concept, confusion comes about because it get mangled beyond belief.
    Some of this mangling is actually physical i.e. we change the shape of a heated object, and other mangling is with words i.e. not knowing much about the subject.
     
    I like to stick with the physical, it is easy and, more importantly, you can put descriptive units to it.
     
    To start with, let us look at the two types of heat capacity.
     
    Specific Heat Capacity
     
    Specific Heat Capacity (cp, SHC) is based on the mass of the material and how much energy (J, joule) it takes to change the temperature by 1 K (kelvin).
    This make for very eas-3y to understand units, but can cause confusion because it is often assumed that the temperature, K, is the important part of the formula and has to be what governs the amount of energy that all materials can store.
     
    J.kg-1.K-1
     
    Different materials have different SHCs and at first sight do not seem to make much sense.
    Air has a higher SHC than granite, and water can experience 3 different SHC in one day.
     
    Air = 1.012 kJ.kg-1.K-1 
    Granite = 7.10 kJ.kg-1.K-1 
     
    Timbers can be very confusing
     
    White Pine = 2.5 kJ.kg-1.K-1 
    Oak = 2 kJ.kg-1.K-1 
    Balsa = 2.9 kJ.kg-1.K-1 
     
    Water(steam 372K) = 2.03 kJ.kg-1.K-1 
    Water (liquid 298K) = 4.18 kJ.kg-1.K-1 
    Water(ice 263K) = 2.05 kJ.kg-1.K-1 
     
    The reason that water, and most materials for that matter, has 3 different SHC is because it changes phase and during that change, the energy absorbed or released can be huge.
     
     
    Volumetric Heat 
     
    This is based on the volume and is usually the more useful unit (cv, cpv, VHC) to use and has the formula
     
    J.m-3.K-1 
     
    It is just the product of the SHC and material density (ρ, Rho, D).  Density is the mass of a material by unit volume, kg.m-3 
     
    Taking typical London brick as an example.
     
    c = cp x ρ 
     
    cpv(brick) = 0.84 [kJ.kg-1.K-1] x 1845 [kg.m-3]
     
    c = 1,550 kJ.m-3.K-1 
     
    As an ordinary London Brick is 0.215m x 0.103m x 0.065m and ignoring the mortar mix (cp = 0.96 kJ.kg-1.K-1, ρ = 2080 kg.m-3), an area of 1 m2 will take approximately 155 kJ to change in temperature by 1 K.
    Depending on the starting conditions i.e. the outside and inside air temperatures, the mean starting temperature of the brick, the thermal conductivity and the time taken to change, the thermal conditions change that 155 kJ number because of the thermal conduction losses.
     
    That is it really, the main things to remember, to save confusion, is that one must specify which heat capacity is being used i.e. by mass or volume, are any phase changes happening i.e. plaster drying out, and make sure the units are shown.
  6. SteamyTea

    Insulation is really easy

    By SteamyTea,


    Insulation
    Why do people get so hung up about thermal insulation, it really is not difficult.
     
    The main thing to remember is that the power, in watts (W, J.s-1) that passes though a material is approximately proportional to three things, conductivity (k, λ), temperature difference (∆T) in kelvin (K), and thickness in metres (m).
    In arithmetic terms, the thermal conductivity of a material is written as W.m-1.K-1, or W/m.K (as I have never found a way to write a superscript negative sign on my Android phone).
    Different materials have different thermal conductivity.  Taking extreme ends of the spectrum, natural diamond conducts at a rate if 2200 W.m-1.K-1 and a pure vacuum, for these purposes, is 0 W.m-1.K-1.
    Now we don't, in the real world, work at the extremes.
    So let us stick to some more common building materials.
    Ordinary brick, k = 0.72 W.m-1.K-1. Concrete k = 1.28 W.m-1.K-1. Timber k = 0.14 W.m-1.K-1. Mineral wool insulation k = 0.038 W.m-1.K-1. Expanded polystyrene k = 0.04 W.m-1.K-1. Polyurethane foam k = 0.03 W.m-1.K-1. Orientated Stand Board (OSB), 6% adhesive k = 0.16 W.m-1.K-1. Plasterboard k = 0.19 W.m-1.K-1.
    There is, obviously a lot more materials and it is down to whoever is calculating to find and check figures.  An example of this is granite, there are a lot of different types and the k-Value can range from 1.73 to 3.98 W.m-1.K-1.  So do your research.
     
    It is not normal to fit a metre thickness of any insulating material, If we did, none of this write up would be necessary.
    Because we use fractional dimensions i.e. 0.2m when planning the insulation levels of a building, the more common thermal resistance (R = K.m-2.W-1) is used.  There are two advantages of using the R-Value, it takes the thickness of the material, and the area of the material, into account.  Converting from the k-Value to the R-Value is really easy, just divide the thickness by the k-Value.
     
    R = l / k
     
    R-Value is often quoted and one thing to be careful of is that imperial units are often quoted.
     
    You may have noticed that R-Value has somehow introduced a m-2 unit, this comes about from dimensional analysis of all the International System of Units (SI) units W, which is kg.m2.s-3 and some arithmetic rearranging when combined with the other units, m and K. This is a useful as we do not have building elements that only have thickness, they also have area.
    The most useful thing about R-Values is that they can be added together to give a total thermal resistance (ΣR).
     
    Taking a simple wall build up of:)
    Outer: Brick, k = 0.72, 0.1m thickness.
    Full Fill Mineral Wool, k = 0.038, 0.2m thickness.
    Inner: Brick, k = 0.72, 0.1m thickness.
    Plasterboard, k = 0.19, 0.012m thickness.
     
    The overall thickness is 0.412m (dimensions may vary, so check).
     
     
    Using the sum (Σ) of l / k for every component makes for a long equation, and it is usual to use a spreadsheet.
     
    ΣR = 0.1 / 0.72(outer brick) + 0.2 / 0.038(mineral wool) + 0.1 / 0.72(inner brick) + 0.012 / 0.19(plaster board)
     
    ΣR = 0.139(outer brick) + 5.263(mineral wool) + 0.139(inner brick) + 0.063(plaster board)
     
    ΣR = 5.604 K.m-2.W-1.  Note here that the effects of the mineral wool are dominant and that large R-Values are better.
     
    It is not normal to talk about a house, or wall, having an R-Value, but a U-Value (anyone know if R should be proceeded with 'a' or 'an', sounds like it should be 'an', but U sounds better with 'a').
     
    Changing to U-Value, which is W.m-2.K-1 is simply a matter of taking the inverse of the R-Value K.m-2.W-1
     
    U-Value = 1 / R
     
    So in this example:
     
    U = 1 / 5.604
     
    U = 0.178 W.m-2.K-1.
     
    If the wall, ceiling, roof or floor is of timber construction, the technique is just the same, just that the appropriate areas also have to be included in the final solution, so you work out the U-Value for all the studs and noggins, plus the OSB thickness, and then the U-Value for all the insulation and the OSB thickness.
     
    There is one other thing when looking at heat losses, and that is the air film surrounding them.  Air has a very good k-Value of 0.026 W.m-1.K-1 and is really the component that is doing the majority of the work in insulation, the material i.e. mineral wool or polyurethane foam is just there to stop the air conducting by trapping it in place.  Because of this, some allowance has to be made for any air voids in the wall build up i.e. a service gap.
    To simplify this, it has been decided that two standard values are used, one for walls and one for roofs, with no regard to thickness.  The wall R-Value is 0.18 K.m-2.W-1, roof R-Value 0.04 K.m-2.W-1.
     
    So taking the above example, and extra 0.18 K.m-2.W-1 must be added to the sum of the R-Value.
     
    ΣR = 5.604 + 0.18
     
    ΣR = 5.784
     
    Convert to U-Value
     
    U = 1 / 5.784
     
    U= 0.173 W.m-2.K-1.
     
    It only makes a small difference, and at the third decimal place, but is still worth including because when the numbers are rounded, it may be the difference between the desired value or not.
     
     
     
  7. SteamyTea

    Insulating Panels; Part 2

    By SteamyTea,


    Thermal Testing
    It has been a lot warmer this week, which is nice, but not so great for doing comparison testing.  While my house stayed within half a degree (17.4°C instead of 17°C) the outside air temperature was 8.5°C warmer (8.3°C compared to -0.1°C) over the test period (20/01/2024 to 24/04/2024).
    But no matter, I still run the tests with just an air void and the void filled with a product called Fillite, which is hollow microspheres made from silica.
     
    Basically the same pattern was observed with the two samples performing in a similar manner.  The temperature differences, within the samples, are different with the air void showing a mean ∆T of 1.1°C and the Fillite at 0.7°C.
    The slope of the trendlines was less, but not significantly so.
     

     
    The slope of the Air line is 0.1, which means that for every 1°C change in the temperature difference in the overall temperature, the sample only changed by 0.1°C
    The Fillite was 0.15, so better.
    The Clingfilm slope was 0.14, so almost as good as the Fillite.
    The Aluminium Foil was 0.2, so the best and twice as good as just an air gap.
     
    So there we have it, possibly.  A bit of tin foil creates a greater temperature difference over the temperatures samples, even of the overall temperature differences do not see so different.
    A small change, of a small amount, equates to a very small change.
  8. SteamyTea

    Insulation Panels

    By SteamyTea,


    Thermal Testing
    It has been a cold week, so just after 12:30 on the 15th of January 2024, I started a small, but limited, experiment.
    This was rather prompted by a comment by @Garald who wanted to insulate at the back of his book shelves, and mentioned our favourite insulation, multifoil.  @Gus Potter also has a project that may benefit a thin, easy and cheap to make, insulating panel.  I think I also made a comment to @saveasteading about this experiment, but can't remember in what context.
    Now I have always been dubious of reflective type insulation.  Works great at high temperature, especially in a vacuum, it is how the cameras on the James Webb Telescope are kept cool.
    But we are not Billionaires, so I used hardboard, pine, white emulsion paint, double sided tape, small screws, clingfilm and aluminium foil.
    Basically I made some small St. Ives picture frames, put them face to face, separating the 24mm air gap with clingfilm in one, and aluminium foil in the other.  The total thickness of the test panels is 30mm.
     

     
    Each side of the test material had temperature sensors (DS18B20s) inserted via holes in the frame.  These had been calibrated before hand and the analysis is based on the calibrated data.
    Other sensors where fitted in the room and externally to log ambient temperatures (why being a cold week was so good).
    The panels were then stuck to my kitchen window with double sided tape.
     

     
    The position of the sensors allows for a combination of temperature differences to be logged, logging was at the minute interval but the analysis was based on 6 minute means.
    A quick calculation to check the standard error showed that accuracy was a factor of at least 10 below the 0.1°C accuracy of the experiment.
     
    The data analysis was based around temperature differences, but for some context, internal and external air temperature is also shown on the charts (right y-axis).  A frequency distribution line was also added, this is black line (right y-axis) and is called Ambient ∆T Probability Percentage.  The Ambient ∆T is the difference between the inside temperature and outside temperature.  This is also used for the 0.1°C temperature bins that create the x-axis.
    Mean temperature differences between each side of the clingfilm or foil (shiny side towards warmer room) were also calculated and binned according to when they happened with respect to the Ambient ∆T.
    This method is used as it is more relevant than a time series that can fluctuate during the day, it is the properties of the insulation that is being tested, not the absolute 'comfort' levels.
     
    The below chart shows the total test period results. Test period (15/01/2024 12:26 to 19/01/2024 12:42)
     

     
    Always remember that these are temperature differences and not absolute temperatures, except the Internal and External mean temperatures (yellow and green lines) and the probabilities (black line).  Those 3 are read from the right hand axis.
    The Clingfilm is the red dots and the Aluminium Foil is the blue dots.  Linear trend lines have been added more for clarity than actual predictions.
    A can be clearly seen, there is not much difference between the two datasets. The Clingfilm performs better overall with a mean difference of 2.5°C, to the Aluminium Foil's 2.3°C.
    Above an Ambient ∆T of 19°C the Aluminium foil performed a little better.  This is actually saying, the colder it is outside, the Aluminium Foil performed better, which may be important comfort, but overall, there will be greater energy losses than with just using Clingfilm as a separator.
    It is, purely from a climate change viewpoint, the overall energy reduction that is important.  Climate change has caused the mean temperature at my end of Cornwall to average -0.1°C for 4 days.  I have lived back here for 20 years and never known such a prolonged cold period.
     
    I am now running a second test, using one panel without any separator at all, and the other one fully filled with silica micro balloons.  Micro balloons may sound exotic, but they are just filler used in the plastics industry.  I don't think the temperature differences are going to be so great next week, which is a shame as the greater the range that can be tested, the better.  I can always raise the temperature in the room to compensate, but as I raised it up to 24.5°C a few times, which resulted in only getting a mean of 23.5°C in the room, it will be a bit costly and not very environmental.
    I shall post up the results of the second test next week, all going well.
     
     
  9. SteamyTea

    Chart Time: Hourly Temperature decay and power inputs.

    By SteamyTea,


    Hourly Temperature Decay and Power usage: Statistics Heavy
    Prompted by @haddock's query here:
     
    and my few charts to show what has happened in my house, I have finally got all my data together and after looking at dozens of charts, have reduced it to two that show the most useful information about my house cooling, or heating.
    Initially thinking that the difference between internal and external temperatures was the best base to chart against, I soon realised that it shows hard to understand results i.e. a larger number, the colder it is.  Then it struck me that as I was looking at the slope of the data points for °C/hour change in internal temperature, temperature difference was built in i.e. colder outside, the faster the house cools.  Real scientists would be talking about 'energy forcing' which cover all energy inputs because of the Conservation of Energy: Energy cannot be created, only change its form.
    But enough of that, the rest is statistics.
    Without going into too much detail about data error checking, rounding and discarding data, the data points for the last 4 years were reduced to about 500,000 from about 300,000,000.  Or basically data collected every 6 seconds reduced to data for every hour.
    The data eventually used was Year, Month, Hour, External Temperature, Internal Temperature, Power and Zero Power.  Other fields were created i.e. Week Number and Maximum Power, but these remain unused at the moment.
    From that data it was quite easy to create hourly temperature slopes [°C/h], just take the last hours temperature away from the current hours temperature.  If it is positive, it has got warmer, negative, colder.  Power data was averaged (mean) over each hour.
    If there was an error because of missing data, or misreported data, then that was filtered out and will show up in a lower data point count but can be corrected with the Standard Error of the Mean.  This left 139,165 usable data point out of 140,268, not perfect, but quite usable.
    There was one other set of derived data that was discarded, and this was extreme temperature slope values.  Generally, the majority of the data points for the temperature slope was between -0.4°C/h and 0.4°C/h.  Occasionally a slope that was into the major integers appeared.  Now it is very possible to get cooling of 2°C/hour by leaving the window open, similarly a large increase in room temperature could be caused by letting a cake cool under the temperature data collector (I actually did this and wondered why the room temperature was reported as 27°C, in November).  This data was therefore filtered to exclude anything outside of the range of >=-0.5°C/h and <=0.5°C/h.  Again, this can be justified as the Sample Count and the Standard Error can be used to adjust the results.
    Other descriptive statistics were used to help explain what is happening.  Minimums, Maximums, Standard Deviation, Skew and Slope were calculated on the relevant data ranges.  External Temperature Distribution, as a percentage, was calculated as a Normal Distribution as well as empirically from the data.  This was done as a data check but also highlights the variations in the temperatures and the associated skew i.e. long tail to the left [negative skew], the modelled data shows no skew.
     
    Now that the dull bit is over, some charts.
    The first chart is all months and all hours for the years 2019, 2020, 2021 and 2022.
     

     
    As expected, the mean slope, the actual change in rate of the house warming up and cooling down, is very close to 0°C/h.  It also shows that at very low external temperatures, the house cools faster i.e. -0.3°C/hour when it is -3°C outside.  Around the mean external temperature of 12°C there is house cooling of -0.06°C/h, which is basically no change.  The house does not start to warm up, until the external temperature is 23°C, but as the Temperature Distributions show, that does not happen very often down here in Cornwall, less than 1% of the time after rounding.
    This does not mean that the house does not get hot, at one stage, it was at 29.43°C, a proper temperature and one I can easily get used.  Taking the internal temperature standard deviation into account, 90% of the time, the house temperature is 20°C ±4°C.  Lower temperatures were probably when I was away, higher temperatures were probably during the two severe heat warnings we have had down here.
     
    Power usage, which is a bit peculiar in a house with storage heaters, often show a warmer temperature in the mornings and it being cooler in the evenings.  As the data can be filtered by hour, the next three charts will show all times over the last four years, then after the heating has finished during the first two months of the year, up to near enough local noon, then local 14:00 to 19:00.
    The x-axis is Internal Temperature Slope, the °C/h.
     

     
    The above chart shows that as the internal temperature change reduces, less overall energy is needed to keep the house a a stable temperature. This is backed up by the times the house is using Zero Energy.  The Zero Power outlier on the left corresponds with an almost zero energy usage, and a greater temperature drop of just over -0.3°C/h, so the house was probably empty then.
     
    The 7AM up to 1PM [6 hours] chart during January and February which are known heating months.
     

     
    This shows a very different picture.  The mean power is now 0.1 kW [100W], down from 0.5 kW [500W] because there is no heating input, but the house is hardly changing in temperature with most of the data points clustered between  -0.02 to 0.05°C/h.  The outliers on the right are probably oven usage.
     
    The 2PM to 7PM [6 hours] chart for the same period.
     

     
    As the storage heaters have now not been recharged for at least 8 Hours and up to 14 hours, it is unsurprising that the house is starting to cool a bit more, with the energy inputs stretched out a bit more between -0.28 to 0.2°C/h, clustering between -0.18 to 0.07°C/h.  This is highlighted better on the next two charts which cover the same time periods.
     
    7AM to 1PM [6 hours] chart.
     

     
    As the temperature slope is negative, it shows that the house is cooling at a rate of -0.003°C/h, which is basically stable and shows that the storage heaters are keeping the place warm, 19.8°C, with a greater warming affect the colder the outside temperature is.
     
    The 2PM to 7PM [6 hours] chart
     

     
    As mentioned above, because the storage heaters have not been charged up for many hours [between 8 and 14 hours], it is not a surprise that the house is cooling.
    What is a surprise is that the relative cooling is so low at 0.015°C/h.  Part of this will be because the internal temperature has risen slightly to 20°C [my target temperature].  Allowing for the increase in the Standard Deviation of 1.32°C, up from 1.29°C in the morning, there is, in reality, no temperature change, and certainty not one that is noticeable.
     
    Now that is all over, I am going to show four charts that highlight what adding secondary glazing and fixing the leaky back door have done.
    These are for December.  The first two are the temperature slope, second two are the power usage.
     
    December 2019, 2020, 2021 Temperature Slope chart.
     

     
    December 2022 Temperature Slope chart.
     

     
    Pre improvements was cooler in the house even though the mean external temperature was 8.3°C as opposed to 5.8°C in December 2022.  The range of temperatures where much greater as well in December 2022.
     
    December 2019, 2020, 2021 Power Usage chart.
     

     
     
    December 2022 Power Usage chart.
     

     
    Mean power is now 0.85 kW, down from 1.28kW, a reduction of 0.43 kW or 320 kWh for the month.
    The slope, on these power charts, shows the change in power needed for every °C change in external temperature.
    That has gone from 100W/°C to 60W/°C.
    Airtightness and insulation really work, and I am still getting the benefit of those improvements.
  10. SteamyTea

    The energy meter experiment

    By SteamyTea,


    Energy meter
    I am not sure how well it works yet, but it works in trivial cases i.e. a 40 W lightbulb and my fridge (once I had stopped stray morning light).  I am going to ask my neighbours if I can pop the Photo Diode on their meter as I don't want to stop the one that is already logging mine.
    What would be good is if others, who have a lot more knowledge and skills than me (I am really just a chancer than fiddles about till it seems to work) could improve and add to it.
    Things that would be nice are a remote sensor to save having to run a small bit of wire into the meter box.  A nicely made sensor cover that holds the magnet in place, and does not let stray light in.  A display, and an enclosure for the complete kit (thinking of you @Onoff and your plastic printing skills).
    Remote, but not cloud based access may be interesting, but as will all new toys, after a few days, does not get used often.
    Other 'things' could easily be added i.e. temperature, humidity, air pressure, but they are really stand alone items, though inside and outside temperature is useful.
     
    So here it is, my feeble attempt.
    What I have used to make my energy meter.
     
    A shop bought energy meter

    https://www.ebay.co.uk/itm/354118466147
    I used this purely to test the power, 1 pulse per Wh, same as my main meter.  I will, as I bought 3 of them, use them on my 3 ‘night circuits’ once I have finished playing.
     
    Some magnets with holes in the middle
     
    https://www.ebay.co.uk/itm/184928442876
    These are to hold the light dependant LED onto the meter.  A suitably sized metal washer was super glued in place.
     
    Some photodiodes to sense the red flashing light on the meter.

     
    https://www.ebay.co.uk/itm/232690475106
    This is the real magic, just wired in between a GPIO pin, 22 in my case and ground.  The important thing is to wire it in the ‘wrong way around’.  So the anode, the longer leg on the diode, is wired to ground, and the cathode, the short leg, is wired to pin 22.
     
    A Raspberry Pi ZeroW 
     
    https://thepihut.com/products/raspberry-pi-zero-w
    Just a bog standard RPi ZeroW
     
    A header

     https://thepihut.com/products/colour-coded-gpio-headers
    Are useful, and I think you can buy the RPi ZeroW with them already soldered in place.  You need it to easily fit the RTC on.
     
    A Real Time Clock

    https://www.ebay.co.uk/itm/234603677979
    While not necessary, I always fit an RTC (real time clock) as I cannot guarantee an internet connection all the time.  You have to muck about with the /boot/config.txt file to include the line
     
    dtoverlay=i2c-rtc,ds3231
     
    and edit the /lib/udev/hwclock-set file to disable the settings with the # symbol
     
    # if [ -e /run/system/system ] ; then
    #    exit 0
    #fi
     
    A USB to TTL Serial adaptor
     
     https://www.ebay.co.uk/itm/203604196200
    Useful when setting up a ‘headless’ RPi.  Just make sure to change the /boot/config.txt to inclide the line
    enable_uart=1
     
     
    The Code
    The code I have used uses Python3 and standard libraries.
    After much searching and thinking, I found that GPIOZero library was quite useful (https://gpiozero.readthedocs.io/en/stable/api_input.html) as it has some useful code for a ‘button’, or switch to the rest of us.
    I also, included a block of code to create a daily *.csv file that automatically changes the filename every midnight.
    All the code does is sense when the light dependant LED senses light, and when that light stops, it timestamps the daily *.csv file, then wait until more light appears again.
    Simple, 3 lines of main code.
     
    #!/usr/bin/python3
    from gpiozero import Button
    import time, datetime
     
    button = Button(22)
    def sort_time():
                dt = datetime.datetime.utcnow()
                runday = dt.day
                dt.day == runday
                ts = time.time()
                UTC = datetime.datetime.utcfromtimestamp(ts)
                logfile = '/home/pi/monitoring/data/meter-%s-%s-%s.csv' % (dt.day, dt.month, dt.year)
                tfile = open(logfile, "a")
                tfile.write("%s"%UTC + '\n')
                tfile.close
               
    while True:
                button.when_pressed
                button.when_released = sort_time
     
    What could be simpler.
     
    The output is presented like this.
     
    2022-07-09 13:05:38.577239
    2022-07-09 13:05:40.028295
    2022-07-09 13:05:40.374854
    2022-07-09 13:05:50.753515
    2022-07-09 13:05:52.390287
     
    Each timestamp is equal to 1 Wh.
     
    I do all the analysis in a spreadsheet.
     
     
×
×
  • Create New...