New Lives For Old Buildings

AThirty years ago the question asked of a disused factory was how quickly it could be cleared. The question now asked, at least in the planning departments of northern European cities, is what might be put inside it, and the reversal has been driven less by affection for old brickwork than by a single line of arithmetic. Manufacturing the concrete, steel, glass and insulation for a new building accounts for a large share of the emissions that building will ever cause, and that share is spent before anyone moves in. Keeping a frame that already stands therefore looks like an immediate saving, banked at the moment of the decision rather than promised over decades. Several cities have written the reasoning into policy: proposals to demolish must now be accompanied by a statement showing that conversion was examined first, and in Ardwell, where the requirement dates from 2016, roughly a third of applications to demolish are withdrawn once that statement has been prepared. The arithmetic is not wrong. What is questionable is the assumption, rarely stated aloud, that a saving made at the start cannot afterwards be given back.

BThat assumption is what Naomi Cardus set out to test. Working from the utility records of 208 conversions completed between 2009 and 2018, her group at the Fenwick Institute of Building Economy compared each project with the new building that had been proposed for the same site and rejected. The conversions had indeed avoided, on average, 340 kilograms of emissions for every square metre of floor area. They then used more energy every winter: the median converted building drew 38 per cent more heat per square metre than the design figures for its rejected rival, and in 38 of the 208 cases the excess was more than double. Cardus calculated a crossover, the year at which the accumulated heating deficit cancels the saving made at construction, and put it at between twenty-three and forty-one years for the sample as a whole. Buildings, she observes drily, are expected to last longer than that. Her conclusion is not that conversion is wasteful but that it is a loan rather than a gift, and that the repayment schedule depends on what is done to the fabric afterwards.

CStefan Voigt, an energy analyst at the Halberg Institute, is sceptical of the comparison itself. The rejected new buildings in the sample were never actually built, so the figures set against real meter readings were design predictions, and design predictions, he maintains, are systematically optimistic. Where new buildings have been monitored after occupation, they have consumed between a quarter and three-fifths more energy than their models promised, a discrepancy the industry calls the performance gap. Correct for it, Voigt argues, and the crossover recedes so far into the future that it stops being a planning consideration. His own measurements, however, cover twelve office buildings through a single heating season in one temperate region, and offices are precisely the type in which patterns of occupancy vary most from year to year. He concedes, moreover, that the gap has narrowed since mandatory post-completion testing was introduced, which weakens his correction exactly where it would matter most, namely for the buildings coming forward for approval today. Neither party disputes the meter readings. What they dispute is what those readings should be measured against.

DBeneath the disagreement about carbon lies a more mundane obstacle. A conversion cannot be priced with any confidence until the fabric has been opened up, and the fabric cannot be opened up until the building has been bought, so the risk sits with whoever moves first. Lenders respond by demanding contingencies that a new building would not carry. The trades involved are also scarce: repointing brickwork, splicing a decayed timber floor or working lime plaster are skills that fell out of general use during the decades when demolition was the default, and colleges in Ardwell reported ninety-one applicants for lime and stone courses in 2019 against several thousand for general construction. There is a further asymmetry that is easy to overlook. Insulating a masonry wall from the inside works well provided that the wall is dry and remains so, but the same treatment applied to a wall that takes driving rain can trap moisture behind it, and the resulting decay stays invisible until it is expensive. The result is that reuse is cheapest and safest in exactly the buildings, sturdy, dry and generously proportioned, that were least likely to be pulled down anyway.

EWhether any of this reaches the people already living around the building is a separate question, and the evidence is uncomfortable. Converted mills and warehouses command rents that their districts did not previously support; a survey of eleven such districts found that average residential rents rose by 27 per cent within four years of the first scheme opening, against 9 per cent in comparable districts nearby. Preservation, in other words, preserves the fabric and not the population, and it does so only where the surrounding market is already rising. My own reading of the evidence is that Cardus has identified the real weakness of current policy and Voigt has identified the real weakness of her method, and that both can be met by a rule neither has proposed: approve conversion, but approve it together with the fabric work that closes the heating deficit, and treat the two as a single decision rather than as a permission followed by a hope. That would make reuse dearer at the outset. It would also stop the profession claiming a saving in the year of construction that the building spends the next four decades handing back.