You might remember that some weeks ago we launched a contest in IESD [the Institute of Energy and Sustainable Development, based in Queens Building here at DMU] to see which one between the Hydro Hot Tap and the Kettle was more efficient for preparing tea and coffee in the office.
I would like to share today the outcome of the contest [also know as The Epic Battle] and to do so I will have a guest blogger in the person of Richard Snape, a PhD researcher in IESD, part of the `engineering team` [also composed by Dr. Paul Cropper, Dr. Michael Coleman, and Dr. Graeme Stuart] that helped me analysing and making sense of the data we have been collecting. I am now leaving the `pen` to him!
The results of the Kettle vs. Hydrotap contest are in (courtesy of Dr. Paul Cropper) and have given us some food for thought (excuse the kitchen-based pun). The data are here [Kettle_vs_hydrotap], the measurements are kWh used in each 1 minute time-slot. Here’s what it looked like:
The graph above makes it look like a hands-down win for the hydrotap! An initial look at the data and simply summing the energy used in week 1 and week 2 shows that significantly less energy was used to heat water in the hydrotap week than the week with the kettle (see table below). However, that’s not the whole story…
The experiment, as announced by Monica in a previous post, was set up to simply measure the energy use associated with the hydrotap and kettle over two weeks, with users asked to use the hydrotap exclusively in the first week and the kettle the second. To encourage this, the kettle was removed from the kitchen for the first week and the hydrotap was switched off at the main feed for the second. The weeks were arbitrarily chosen and a number of people raised the issue that there may be a greater number of people in the office (and therefore greater demand for hot water) in one week than the other. This was particularly pertinent as one of the weeks chosen happened to coincide with the half term holidays in local schools.
This got us thinking about how we might do some further analysis on the data to work out how often the devices were used. Firstly, I filtered the data to look for heating “events” – defined as consumption changing from low power “normal” state to “in use” state (>0kWh/min for kettle, >0.004kWh/min for hydrotap – an assumption based on observation of the data). This method found that in the hydrotap week there were 159 events and in the kettle week there were 189 events. Aha – the kettle was used more often, maybe that’s why it used more energy…
To test this, I worked out the average energy per event as well – again shown in the table below
|Total energy in week (kWh)||Total heating events in week||Average energy per heating event(kWh)|
Still, the hydrotap wins by a fairly convincing margin.
Analysis didn’t stop there, however. Dr. Graeme Stuart was interested in how the hydrotap in particular used its total energy. To that end, he calculated the average power of the device in each 1 minute slot and then a histogram of number of 1 minute timeslots at each power level. The results for the hydrotap are shown below. With the kettle, we know that the kettle will be on and consuming at a rate of around 2.5kW when it’s heating. The spread in value of energy used in each minute slot is simply down to the fraction of each minute the kettle is on. The hydrotap is more complicated. There are a large number of minutes with very low consumption, then a sort of spread out block of slightly larger consumption and then a peak of relatively high consumption(“full power”). When these are plotted cumulatively, we see roughly ¼ of the hydrotap energy consumed at very low power, ¼ at medium power ¼ between medium and full power (probably fractions of a minute at full power) and the rest at “full power”. We hypothesise that the low power consumption is the unit in genuine standby (e.g. at weekends and overnight), the medium power is keeping the water up to temperature in the storage tank and the high power is heating after someone actually uses some hot water and cold water enters the tank. This needs a bit more thought to draw any true conclusions, but indicates that switching the hydrotap off overnight and weekends might save around ¼ of its already lower consumption.
I was still intrigued by the number of “events” and, indeed, why the consumption of the kettle was so much worse even though the hydrotap was keeping a 2 litre tank of water warm every day. So I took each “event” for the kettle and looked at the total energy used between it starting and stopping. I then approximated the volume of water heated (another set of assumptions – based on heating from cold – 12oC – to boiling – 100 oC, SHC=4200 J/kg.K). If that approximated volume is less than 150 ml (another assumption – but I reckon it’s hard to actually boil less than 150ml in a kettle), I assume it is a “reheat” of the previous boil. This shows that 30 of the 189 kettle events are likely to be “reheats” i.e. where people have boiled the kettle and then they (or someone else) have reheated it shortly afterwards. Of the non-reheat events, the approximate average volume of water heated is between 0.5 and 0.6 litres. If we assume that most people want a single cup of water, this means they heat between 2 and 3 times the volume of water they need.
There is more that can still be done. We could work out a mathematical model of each device and look at the main parameters affecting energy use – the factor by which people heat more water than they need in a kettle and the frequency of use. This would give some thresholds for where a hydrotap uses less energy than a kettle or vice versa. We could also find what size of hydrotap installation would be most efficient for a given usage scenario. We could conduct the experiment again asking people to note what they were using the water for to get a better estimate of how much water they actually need vs. how much they use, although we might suspect such an experimental setup would have a fairly big impact on their behaviour.
The main conclusion, though, is that the hydrotap uses less energy than the kettle and is more efficient even for our fairly low usage frequency (~30 uses per day). It’s amazing how much you can work out from a single time series of energy consumption data – thanks for prompting this interesting ad hoc experiment and analysis!
J Richard Snape, PhD Student
Thanks so much to Richard for writing this post; to Paul for being so helpful in setting up the data collection; to Graeme for the help in analysing the data; to Michael for the enthusiasm and the help in setting up the experiment and analysing the data; to all the IESD staff and students for the willingness to help and their precious collaboration during the two weeks experiment!
What do you think? Did you like our contest? Are you satisfied with the results? Have you got more questions? Post them in the comments section and we will ask our experts!
For me, the experience was really fun! Thanks everybody! Happy to have transformed IESD, and DMU, in a Living Laboratory for a couple of weeks!