Many will have a reasonable understanding of concepts such as efficiency and smoke output, but probably not much idea of how those numbers on data plates and Declaration of Performance certs are actually generated. Contrary to some rumours, the numbers are not just made up. There is a fair degree of science behind them and a phenomenally expensive amount of equipment and testing hours to get an appliance through certification. This article outlines some of the core principles, procedures and assumptions in stove testing.
What is tested and why do we bother?
Heating appliances must be CE/UKCA marked in order to be legally placed on the market. To demonstrate suitability for the CE/UKCA mark, an appliance must go through a standardised testing process to prove it meets minimum requirements. The core CE testing process is split into testing for thermal efficiency and emissions, and safety. There is a minimum requirement for efficiency and maximums for emissions of combustion products i.e. CO2, CO, Organic Gaseous Compounds (Hydrocarbons), Soot and NOX. DEFRA Smoke exemption testing is a different set of tests. Perhaps surprisingly, there is no requirement on spillage or testing for spillage because the majority of appliances are classified as ‘Closed Appliances’. So in normal operation, the firebox doors are to remain closed. For an open fire or an appliance that can operate with open doors, there are requirements for spillage, but these are visual, i.e. is smoke visibly spilling. Actual measurements are not taken.
Definitions
Thermal Efficiency – If we put 100 units of energy into a stove and get 80 back in useful heat, it’s 80% efficient. 60 back would be 60%, etc. The remainder of the energy goes up the flue as hot gases and burned, unburned and partially burned products of combustion. Here is the first quirk of testing an appliance: We don’t actually measure the useful heat at all. We calculate the energy that goes in to the appliance (based on mass of fuel and moisture content) and we measure the waste energy going up the flue. We then assume the heat to the room must be heat in minus heat up the flue. Using the first law of thermodynamics (energy cannot be created or destroyed), this is a pretty reasonable assumption.
We calculate the energy going up the flue using the relatively simple equation Q = m.c.(T2-T1), where Q is Energy, m is mass flow, c is a constant dependent on the constituents of the gas, T2 is the flue gas temperature and T1 is ambient room temperature. As this relates to the “waste” heat, the goal is to minimise Q, we want the smallest number possible to give us the best efficiency. So, we want a small mass flow, a small gas constant and, to a chimney sweeps horror, we want the lowest flue gas temperature possible.
CO2 – There’s no limit on CO2 but measuring it is important in determining Combustion Efficiency (different to Thermal Efficiency). As wood contains carbon, when it is reacted (burned) with oxygen we get CO2 and heat. In an ideal world we would have 100% Combustion Efficiency producing only CO2 and water vapour, but it’s never ideal. The goal is to maximise CO2 output because if we don’t, we get partial combustion and the following unwanted products (pollutants) coming from a chimney.
CO – Carbon Monoxide is an unwanted pollutant. CO is the partial reaction of carbon and oxygen and typically occurs because there has not been enough time or temperature for the full reaction. The ideal amount of CO would be zero but unfortunately this is never achieved.
OGC – Organic Gaseous Compounds are also pollutants resulting incomplete combustion. These are predominantly the longer chain hydrocarbons such as Benzene and its derivatives. The goal for this is also zero.
NOX – Oxides of Nitrogen, mainly NO (Nitric Oxide) occur when Nitrogen is combined with Oxygen at high temperatures, typically in the flame front. High temperature causes the N2 bond to break and it happily joins a nearby Oxygen molecule. If we could maintain the temperature for long enough, the process would tend to reverse but unfortunately flame fronts lose temperature rapidly and as such the NOX is trapped and released in its unwanted state. We don’t want NOX is because it combines with unburned hydrocarbons in the atmosphere and, driven by sunlight, forms smog.
Soot – the most complex product of combustion of all. In theory, quite simple but in practice it is the most difficult to understand and control. Soot is formed by heating hydrocarbons to high temperatures in the absence of Oxygen in a process known as pyrolysis. The high temperatures cause the hydrocarbons to break down but in the absence of oxygen they do not form CO2 and instead clump together to make soot. Despite its looks, it is not pure black carbon. It contains many other compounds, quite literally hundreds. Soot in the gasses is measured by extracting a sample from the flue across a small, heated filter paper. The filter is conditioned and weighed before the test and again after the test. The difference in weight gives us the mass of soot.
The testing process for efficiency and emissions
Now we know what we need to measure, we need to actually measure it. A standardised testing process is used across Europe and the official tests can only be carried out by approved test houses called Notified Bodies. Historically the standard used is BS EN 13240. This is currently being updated to BS/CEN 16510. This standard has been published so some test houses have been applying it since early 2023. Others are not as there is a ‘period of co-existence’ when either standard may be used.
For testing, a specified special flue system is used which contains the tappings for all the appropriate measurements to be taken and for the appropriate flue draft control. Dee diagram below:
The appliance is placed on a platform weighing scales and the test flue section is connected to the outlet and sealed. The top of the flue is connected to an extractor hood which allows for draught control and fresh air to dilute the flue gas (this will be explained later). The flue draught is set to 12Pa. For installers wondering why most stoves have a minimum flue draught of 12Pa, this is the reason. It is written in to the procedure and manufacturers must publish the minimum flue draught requirement. It is permissible to specify and test to a different minimum flue draught, but the majority stick with 12Pa.
Once the test rig is set and all of the equipment has been checked and calibrated, testing can begin. The fire is lit and the stove is warmed up for around 30-40 minutes. It is then fuelled with the first test load for what is effectively a ‘sighting lap’. Usually this is not a full test as the engineer will check all the equipment is working correctly and that the specified fuel load will burn for the correct amount of time, i.e. air control settings are correct.
The fuel used must be Birch, Beech or Hornbeam, other species are not allowed. Beech appears to be the favourite amongst test houses. It is nearly always kiln dried with a moisture content typically in the range of 14-18%. The fuel load is calculated by accounting for the appliance nominal heat output, efficiency, moisture content and calorific value of the wood as per the equation below.
𝑚𝑓𝑢𝑒𝑙=360000×𝑃×𝑡 𝐻𝑖,𝑓×𝜂mfuel=360000×P×t Hi,f×𝜂
Where mfuel is the mass of fuel in kg, P = nominal heat output (kW), t = minimum refuelling time in hours, Hi.f, = lower calorific value of fuel (accounting for moisture), η= minimum appliance efficiency.
Most modern appliances are around 80% efficient so a test fuel load for a 5kW stove is approx 1kg. An 8kW stove will be around 1.3 kg. It’s an interesting little experiment to physically see the log size for these fuel loads and understand just how little fuel is needed to achieve nominal heat outputs.
Back to the test process, there is no limit on the number of ‘sighting lap’ tests that can be conducted to make sure the appliance is operating correctly. There is a slight difference between 13240 and 16510 for minimum test duration. 13240 used weight to determine the end of test. The weighing scales is zeroed just before the fuel is loaded and 45 mins (+- 4 mins) after the fuel is loaded, the weight must be back to within 50 grams of the zero weight.
16510 allows us to get rid of the scales and run the test for at least 40 minutes, there is no maximum duration. The end of test is visually determined to be when the flames have extinguished. This is backed up with recording the CO2 concentration in the flue gas at this time and subsequent tests must also return the same CO2 concentration for the test to be classed as finished.
Once the test is deemed complete, the data logging stops. The filters are changed, new fuel is loaded, and the process begins again. I know many sweeps question why manufacturers tell customers to refuel on embers. This is the reason – it is legislated in the test process to refuel on embers. My personal feeling on this is that it is the correct advice. By the time of refuelling there will be at least a logs worth of charcoal glowing nicely in the bottom of the firebox. This will be creating plenty of draught with no smoke output. If there is still flame in the firebox and the door is opened, there will be smoke (visible flame is glowing particle matter) meaning it is almost guaranteed to have a little smoke spilling back. If the wood is dry, it will ignite within seconds of being placed on embers.
An appliance must pass 3 tests in one day, and 2 of them must be back-to-back. The published test results are the average of these 3. Passing 3 tests might sound relatively trivial, I can assure you it’s anything but. Using a natural product like wood makes each test different. You don’t know what you’re dealing with until the test is under way, and you are no longer allowed to make adjustments to the air settings. I have experienced a test going wonderfully with what looks like record breaking results only for the log to shift in the firebox halfway through and the final test result is a miserable failure. Likewise, I’ve seen 2 tests that appeared to be exactly the same and both pass, but the numbers they generate are 25-30% different to each other and it’s impossible to figure out why. A test which could be borderline pass on soot, and the next time an adjustment of 0.5mm on the air lever is made and then it’s a clear pass with no issues at all. The ecodesign limits and particularly testing to the 16510 standard are a serious technical challenge.
It can take anywhere from 4 – 10 tests to get the 3 passing tests in one day, and that’s just for the efficiency and emissions part of the process. Once that’s been done there is the DEFRA Exemption testing and safety testing. See part 2 in the next issue.
Stove Testing Part 2 coming soon – see the News section on this site
