Tuesday, March 24, 2009

Display energy certificates - a missed opportunity to communicate

UK legislation requires that many large buildings display a certificate, updated every few months, that shows "how efficiently the building is being used".

This mandatory certificate could have been used to communicate information to people, and to engage the building's users in the challenge of improving the building's energy consumption. However, it seems to me that the designers of the certificate have almost completely blown it: the certificate's main features looks fairly colourful, but they convey amazingly close to no information at all.

How much information can be conveyed on a sheet of paper? One of the simplest principles of communication is that the message should depend on something; the message should not be fixed in advance. For example, Lord Nelson, at sea, had a collection of a few dozen flags from which he could select some to run up his mast. Which ones he ran up his mast depended on what message he was intending to communicate. Someone looking at Nelson's ship would not know in advance what the flags would look like.

This communication principle is almost entirely lost from Display energy certificates: the look of the certificate is almost entirely determined and fixed before any building data are collected.

The most prominent feature of the certificate is the "A-to-G" scale, with its green-to-red/brown colour scheme. Almost all the numbers on this scale are fixed; the only adjustable piece is the little arrow that points at "how well this building is being used". This performance is measured in meaningless pseudo-units, with "100" corresponding to "average for buildings of this type". This number can't be compared, from building to building, since two buildings might be of different types, and the certificate doesn't say anything about the type. Nor can an ordinary person work out what the number means, nor what they should do about it, because the number can be computed only by experts using the government-approved software that churns out these certificates. (I've looked on government websites, and have been unable to find any definition of the magic formula for computing the number; I imagine I would have to pay to attend a government-sanctioned course in Display energy certificate cookery.)

The second feature of the certificate is the top-right blue rectangle. Again, this object achieves amazingly little communication. It is meant to show how much CO2 the building's use is emitting. I would like to make a prediction: I predict that, on the first certificates displayed in the year 2009 in all the thousands of buildings across the country, every single certificate will have a blue rectangle of exactly the same height!. I make this prediction because it looks to me as if the government-sanctioned standardized software auto-scales the entire blue rectangle so that it has got a standard size! Therefore the only way to find out the CO2 emissions of the building is to look really closely at the scale of the graph, which shows, in the smallest font conceivable, an absurdly long number, at the top of the vertical axis, partly overlapping the axis. This absurdly long number, I would guess, is the actual CO2 emissions. In my photo I think it shows 26095 tonnes of CO2 per year. If someone ever manages to read this number (please bring a magnifying glass!), will it mean anything to them? Is a typewritten number a good way to convey information? Is it a good idea to show five decimal places of precision? When communicators discuss how to label the axes of a graph, does anyone recommend that the six tics on the graph should be labelled (nothing), (8693), (nothing), (17396), (nothing), and (26095)?

The one interesting fact that is well conveyed by the blue rectangle is the breakdown of CO2 emissions between electricity and heating: the top (light) half of the blue rectangle shows the electricity contribution, and the lower (dark) half shows the heating.

And finally, the third prominent colour element in the display is the "Previous operational rating" graph (bottom right, orange), which shows "how efficiently energy has been used in this building over the last three accounting periods". At present (in early 2009), this colour object conveys no information at all as it shows only a repeat of the energy performance rating displayed on the left-hand side. In due course, maybe it will reveal an interesting trend, but it will depend on the choice of "accounting period". Is the accounting period going to be one year? If so, the comparison of last year with the year before will be meaningful, but is it going to engage users? Imagine if, every day when you entered your building during 2009, the porter informed you what the total energy consumption had been in 2007 and 2008. Would these facts interest you in putting effort into efficiency drives? I fear that a yearly update is too long a timescale for any useful engagement to happen. On the other hand, if the "accounting period" lasts, say three months, then the variation in operational rating from quarter to quarter would be entirely dominated by seasonal effects. My guess is that the certificates will be updated annually, so building-users will become completely blind to the certificate. The opportunity for user engagement is almost entirely lost.

What have we seen so far?

1. the main colourful numbers are in meaningless units. As it says, "the numbers do not represent actual units of energy consumed; they represent comparative energy efficiency. 100 would be typical for this kind of building."
2. the CO2 emissions are displayed in meaningful units, but these numbers are not displayed in a way that an ordinary person can understand.

Does the certificate have any useful information on it? Yes, hidden away in tiny print at the bottom left hand side are interesting numbers - at least to building energy specialists. The "technical information" shows, in a table, the energy use (heating and electrical), expressed in the meaningful units of kWh per square metre per year. And to make this energy use comprehensible, the table also specifies the "typical use" (of buildings "like this", I presume).

What could have been done better? As far as I can tell, the entire piece of paper is really conveying just two numbers:

`this building's heating consumption is: 519 kWh/m2/y';
`this building's electrical consumption is: 249 kWh/m2/y'.

The certificate hides these two numbers in the technical corner, displays some unknown munging of them on to the A-to-G scale, displays their effective carbon-ratio in the blue rectangle, and shows how the total changed compared to earlier years.

How could the certificate be better? Well, it could have been better in two ways:

1. the certificate could display the two numbers that it is meant to communicate more clearly, more accessibly, more meaningfully, and more educationally.
2. the certificate could communicate more than two numbers.

Let me spell out what I mean.

1. the certificate could display the two numbers that it is meant to communicate more clearly, more accessibly, more meaningfully, and more educationally. For example, the energy consumption (per square metre) could be displayed visually on a scale that shows the energy consumptions of a bunch of other real buildings - so as to help people visualize and aspire. Looking at the current A-to-G scale, someone in a "D"-performing building may well ask "does any building like mine ever get an A or even a B?" They don't know if it is at all plausible. If energy consumption were compared with that of benchmark buildings (eg, Whitehall, the Swiss Re tower, Freda's flower shop, Cambridge University Library, BedZed), then people would see what is possible, and could get a message such as "my building is as bad as Whitehall!" or "we're using 20 times as much as BedZed." Given real comparative data, people could aspire to meaningful goals.
2. the certificate could communicate more than two numbers. For example, there could be a duty to display and compare energy consumption every month or every week (for at least some number of consecutive weeks per year). Then the regularly-updated certificate could engage building-users in the challenge of energy-saving. If someone tries an energy-saving action, they need to get feedback within a week to tell them whether it made a difference. Without rapid feedback, no-one will be interested in energy saving ideas.

PS - To see the whole certificate in one image, follow this link


Sam Sandqvist said...

Perhaps it would be helpful if you could sketch a, in your opinion, meaningful certificate.


Stephen Henderson said...

I think the annualisation comparison is sensible; the shortest period without seasonal effect.

The units of calculation are presumably?? something like energy consumption per sq m of space-- or per worker. This should be plotted.

The compariosn to other buildings cannot sensibly be represented but should be registered and an analysis available online.

David Pollard said...

From a political perspective, this is an excellent scheme. The colourful notices proclaim that Something Is Being Done; the training of energy assessors can be promoted as 'job creation'; the number of gauleiters in government employ is increased and wannabe gauleiters are encouraged; the costs are largely borne by the private sector; blame and responsibility is generally placed on someone else; and the over-simplification helps to keep public opinion dumbed down.

Unknown said...

Hmm, yes, I'd like to know what allowances are made for 'that sort of building'.

At home with near continuous occupation 24x7 we have electricity ~25kWh/m^2/y and gas/heat ~100kWh/m^2/y (though I hope to halve the gas/heat demand and we are net electricity exporters, so really -5kWh/m^2/y or something on that front!).

So your example building is astonishingly bad compared to my little ex-council house!

Anyway, maybe this certificate version can serve as a straw man and get fixed in a few years' time?



housingdabble said...

interesting comments. I think you may find that as they are done each year (legislation requires annual renewal of certificates) that teach area will become more relevant. So the scale for example will be off little consequence, but rather a visual improvement on the previous year.
That said, i agree ti does little to change behavoiur of those using the building unless supported by measures to change the usage and habits of the users/controls.

Vilnis said...

Yes, yes, yes, yes, yes! But there is an even more elegant solution and that is to report the "absolute performance deficit" which in this case for gas is (519-249)x209159 = 56,472,930 kWh, and for electricity (44-80)x209159= -7,529,724 kWh (credit).

This is important because public authorities could be breaking their backs to promote small G-rated buildings to Fs for the sake of appearances, when more absolute energy could be saved by taking one large building from the bottom to the top of the E band (to take one hypothetical example). In other words, DECs probably cause misallocation of economic resources.