Housebuilders can specify all the green technology they want, but what happens when human beings get left in charge of the thermostat? Buro Happold installed sensors to find out, then told Thomas Lane what they discovered

When a design team draws up its plans for a low-carbon housing development, how much can it predict about how it will perform after it has been built and occupied? And what are the chances that it will be right?

This is a more difficult question to answer than it may seem to a layman. On the one hand, engineers know a lot about the thermal performance of buildings, and have sophisticated computer programs to model it. On the other hand, technology doesn’t always work as it is intended to, and the people who live in the building do all kinds of things that designers couldn’t begin to predict.

The question concerning technology is illustrated at BedZed, still the country’s best know sustainable housing scheme. Apart from some problems with overheating, the residents speak highly of their homes; but when it comes to the complicated automatic woodchip-fired CHP plant, the smiles vanish. It never worked properly and was abandoned in favour of gas boilers. Bill Dunster, the scheme’s architect, says commissioning the district heating was a “nightmare” and on his later schemes he has specified networks of small biomass boilers supplying six homes each.

The human factor is illustrated by the Clay Fields development in Elmswell, Suffolk. This is a low-carbon social housing scheme of 26 homes completed in September 2008. On the face of it, it is performing pretty much as its designers expected. Buro Happold’s dynamic thermal modelling software predicted that the scheme would use 87kWh/m2/yr of power for heating and hot water. (The less sophisticated SAP program, which is used to measure Part L compliance, produced a figure of 121). The built scheme was found to have a power use of 93, which is 52% better than the UK average for existing homes, which is 191.5. So far, so good.

The design aspiration for total electricity use was 13.9Wh/m2/yr, but residents actually consume 40.9. Although no breakdown is available for lighting and ventilation, Buro Happold estimates that they account for between 11.5 and 16.5, which is close to the design aspiration. Electricity use overall is slightly below the UK average of 47.7.

Where the scheme really scores is on water use: this is an average of 116 litres per person per day, of which 25 come from the rainwater harvesting system.

This means that just 91 litres are drawn from the mains, equivalent to level five of the Code for Sustainable Homes. The UK average is 148.

But there is a caveat: some residents use at least three times as much heat, electricity and water as others. Buro Happold discovered this by monitoring the houses using a mix of sensors and questionnaires. It also carried out a behavioural study to find out why there was such a big difference between the best and worst performing homes.

Welcome to Elmswell

Elmswell is a sort of BedZed Lite: it doesn’t set out to be zero carbon, but has solid green credentials with careful consideration given to materials selection and energy use. The walls of the homes are built from hemcrete, a mix of lime and hemp with good insulation properties. The homes are heated with a biomass community heating scheme.

The homes are pleasant places to live, according to Buro Happold’s comfort satisfaction survey. There is generous glazing that makes the most of natural light, and unlike BedZed, the district heating system works well: if residents turn up their thermostat, the heat comes through quickly.

The most profligate residents used at least three times as much heat, electricity and water than the most frugal

Plentiful glazing and insulation don’t come at the expense of overheating. In the summer, internal temperatures hover around 23.5ºC for a brief period during the day, within most people’s comfort zone. The highest temperature recorded was 28ºC, but this was for less than 15 hours over the whole year and was in the south-facing kitchen of the home using more energy than anyone else. Night temperatures are comfortable at 22ºC.

Mechanical ventilation with heat recovery doesn’t mean stuffy homes. “Nobody mentioned it, which is quite important when you’re doing comfort analysis,” says Zack Gill, the research engineer at Buro Happold who is responsible for the post-occupancy evaluation. This was borne out by testing:

Gill tested the system and found there were 0.3 air changes an hour, compared with a recommended minimum of 0.2. The system has a boost control that lifts the airchange rate to one an hour. Open a window and the rate jumps to 17.

So why do some residents use three times the resources of others?

The lifestyles of others

The most frugal household gets through 46Wh/m2/yr for space heating and hot water, compared with the most profligate, who gets through 145. Electricity use follows the same pattern: the highest user clocks up 64.7Wh/m2/yr compared with 17.6 for the most frugal: 3.6 times less. Water use is even more variable ranging from 196 litres of water per person per day down to just 28 litres of water per person per day, 5.2 times less.

According to Buro Happold there is a direct correlation between the use of resources, and three key factors. These are as follows:

  • What people believe. For example, if they believe that turning off the lights saves electricity, then they are more likely to.
  • What they consider normal. They may believe, for example, that everybody has, or ought to have, a 40-inch flatscreen television and a Playstation 3.
  • How much difference people think their behaviour affects the environment.

The researchers asked people a series of structured questions intended to explore their attitudes to the use of heat, electricity and water. These were turned into numbers and plotted on a graph against resource use; as you might expect, there was a strong correlation between attitudes and the use of heat and electricity.

Gluttons and misers

So how do attitudes translate into behaviour? Take the highest and lowest users of heat. Both households live in three-bedroom houses, both are made up of mothers with children and both have similar occupancy patterns – all of which makes the two cases suitable for comparison. In fact, the only difference was that the gluttons were in an end-of-terrace and the misers were in mid-terrace. This is a significant difference: on average, end-of-terrace homes used 25% more heat than their mid-terrace counterparts. But that wouldn’t, by itself, account for the size of the disparity.

Setting your thermostat ought to be easy, but the resident who recorded the highest energy use struggled. “She knew how to change the temperature settings but didn’t know what they meant,” says Zack Gill, who handled the post-occupancy evaluation for Buro Happold. Monitoring shows that she initially set the thermostat to 27ºC. She gradually reduced the setting, but even so it was set to 24ºC in the mornings, which is warmer than most people have their homes. At night in the winter the thermostat was set to 22º and in the summer 18º. All the thermostatic radiator valves were set to maximum, too. At the same time, Gill says windows were left open: “She really liked the breeze through the house.”

When we talk about houses being zero carbon, this will be almost impossible without talking to the users

Zack Gill, Buro Happold

Contrast this with the low heat user. Her thermostat was set to a chilly 17º all the time and manually boosted when she wanted more heat; she also kept her windows closed. Given that the average hot water use is 20Wh/m2/yr, her total heat use of 46 is very impressive.

Elsewhere the average temperature across the development was 22º, which is high. Gill is concerned this could be down to the “Snackwell effect”, a phenomenon whereby people will eat more biscuits if they are low fat in the misguided belief they won’t put on weight. If the same logic is applied to low-energy homes, it means people will crank up the heating simply because they’re living in a low energy home. “If everyone moves into a low energy home and wants 22º, then we have a problem,” he says.

Understanding low energy technologies

According to Gill, two-thirds of residents were not able to program their thermostats. Good controls are an important part of reducing energy use, so the devices provided on this scheme allow residents to set five different on/off periods a day with different temperatures for each period, if so desired. There are also two different day settings so residents can have one setting for the weekend and another for the week. “Most people knew how to manually override the system and turn it off,” says Gill. “Even though heat consumption was low overall, there is still a lot of waste in there.”

Although the mechanical ventilation with heat recovery worked well, inevitably there are times when people want a bit of extra air. This is the idea behind the boost control on the extractor over the cooker in the kitchen that expels the warm air. But what do people do when they feel stuffy? They throw open the windows. “We found that 95% of people opened the windows when they wanted some air,” says Gill. “Most people didn’t realise they could turn the fan up and the system is recovering the heat.” Indeed Gill reckons this could be a difficult habit to crack. “People really like getting a breeze through the house. It’s a high satisfaction point,” he explains. “If people couldn’t open the windows, satisfaction would really drop.”

Lessons for designers

The good news is that low carbon design has moved on since BedZed. Although there have been teething problems with the biomass boiler, the district heating system functions well and the homes are performing close to design predictions.

We may be getting better at designing buildings that perform near to predictions, but this has had the effect of turning the energy reduction spotlight onto occupant behaviour, an issue that will get bigger as the potential energy use of buildings gets smaller. “The next step of low-energy building will be a whole different kettle of fish. When we talk about houses being zero carbon, this will be almost impossible without talking to the users,” says Gill.

What can be done about this? Gill doesn’t think designers can do much apart from tweaks such as making heating programs simpler to use. Ultimately it will be a case of educating users and making them aware of how their behaviour affects resource use. “We do need to understand what help and what motivations we can give to occupants, and what we can learn from frugal users to facilitate lower energy use,” says Gill.

This could include smart meters in homes showing how much water, heat and electricity occupiers are using, measured in money rather than units. Gill says people at Clay Fields wanted to know how much energy other people were using so perhaps league tables of energy use could focus minds. This isn’t as outlandish as it sounds as league tables are a key part of the carbon reduction commitment and are a powerful driver in modifying behaviour. The next stage of Gill’s project is to educate people about energy use at Clay Fields to see if this helps drive down energy use. It will be interesting to see if you really can get people to turn the lights off.

The development

The 26-home scheme was designed by architect Riches Hawley Mikhail for Orwell Housing Association. Located in the village of Elmswell in Suffolk, it aimed to be an exemplar for providing good quality, high-density, low impact, inexpensive housing for people in rural communities. It includes nine three-bedroom and 13 two-bedroom houses and four one-bedroom flats arranged in blocks of three, which helps with the high-density aspirations and cuts energy use.

The homes are arranged in four groups of two rows, each of which has three houses. There is an orchard, allotments and public spaces in between. Each row is orientated north–south to maximise solar gain. Each group of homes is offset from the others and the southernmost row is lower than the other to minimise solar shading.

The homes are timber framed with walls built from hemcrete, a mix of hemp and lime, which has excellent insulation properties and minimal environmental impact. It is finished in lime render and western red cedar shingles. The hemcrete is sprayed on to improve airtightness, an approach vindicated as the air leakage test is just 3.5m3/m2/hr at 50Pa which is better than the target of 5 and a significant improvement on the Building Regulations’ requirement of 10. Mechanical ventilation with heat recovery minimises ventilation heat losses. Rooflights are used to improve airflow through the homes in the summer.

The scheme is heated with a biomass boiler sized to provide 70% of the total load with the balance provided using gas boilers. The heat is circulated through insulated pipes via a heat exchanger in each house. Heat meters are used as the basis for charging residents for space heating and hot water use.

Water use

Water use was the wild card, with huge variation between high and low users – by a factor of seven. Unlike with heat and electricity, there was no correlation between attitudes and water. On the positive side overall water use, the mains element of which is metered, is very low and the rainwater harvesting system is seen as a positive feature. “It’s a slightly different colour which people were cautious about at first but now they like it because they know it’s saving them money,” Gill says.

The biomass boiler

The only element of the scheme that hasn’t come up to scratch is the biomass boiler. Unlike BedZed’s CHP system, this just burns chips or pellets for heat but there have still been problems. Over the monitoring period it supplied just 42% of the heat rather than the design target of 70% – the rest has come from the top-up gas boilers. Gill says the biomass boiler wasn’t used for the first three months as it was awaiting commissioning. The moving floor in the pellet store that moves the pellets towards the boiler supply auger failed for a while and it took time to set up a maintenance contract for the boiler. These issues have been ironed out now and at the end of the monitoring period the biomass utilisation rate had jumped to 59%.

Electricity use

Like heating, electricity use varied considerably across the development. The highest user gets through 64.7Wh/m2/yr compared with 17.6 for the lowest user, 3.6 times less. Interestingly, although there was a correlation between attitudes and high electricity use, the people who used a lot of heat didn’t necessarily use a lot of electricity. Gill thinks the lighting and ventilation electricity loads are quite low, with appliances adding up to 65% of energy use. He doesn’t have data for high electricity use but puts it down to energy-guzzling devices such as tumble dryers. Although this element of energy use is nothing to do with the design of the home, it would be an issue for designers of zero-carbon homes, particularly as the biomass heating system means 62% of carbon emissions comes from electricity use.