Engineers Without Borders challenge

The brief

Engineers Without Borders UK’s Engineering for People Design Challenge asks student teams to design for a real community. Ours was Govan in Glasgow, on the built environment and energy theme. Many of its older homes have high ceilings, thick, poorly insulated walls, leaks and damp, and nearly a quarter of households in the wider Glasgow region are in fuel poverty. The solution had to be cost-effective, deal with leaks and damp, and make home heating more efficient.

My part

Early concepts

I proposed two concepts, one of them modular housing, which wasn’t feasible for a large population. The team went with a teammate’s idea of heating skirting boards, and I switched to supporting it.

Context research

My task on the team plan (2–21 November 2024) was researching Govan’s problems and possible solutions from social, economic and ergonomic angles. It fed into the script for our presentation.

Materials

I researched materials that balanced affordability, sustainability and thermal conductivity, and argued for copper and aluminium alloys.

Presentation

My part covered cleaning and maintenance: unlike radiators, skirting heaters don’t spread dust around a room, which suits hospitals and public buildings such as Govan’s library.

The design

We proposed skirting-board heating, using the commercial ThermaSkirt system: an aluminium skirting panel carrying two or three oval copper–aluminium alloy pipes, fed by a boiler or heat pump. Water-fed versions heat more slowly than electric ones but hold their heat for longer.

(opens full size)

A weighted decision matrix compared four options on service level, installation, cost, life cycle and ease of termination:

Looking back

Losing my own concepts early was a useful lesson in adapting and backing a better idea. But the engineering case we made was thinner than the poster suggests.

• The matrix’s weighting was inverted. Its key says 1 is the most important weighting, but each total multiplies score by weight, so “ease of termination”, labelled least important, counted four times as much as “service level”. Re-running it with the weights the right way round, or with none, keeps the same ranking, but the lead over insulation is small in every version. One cell was also mis-multiplied: radiant panels should score 29.
• Insulation isn’t an alternative to a heater. It reduces the heat any heater has to supply, so for poorly insulated tenements the fair comparison is insulation plus a heat source.
• The heat-transfer model didn’t describe the product. Our conduction estimate sent heat 2 m along a solid metal bar from boiler to panel. In a wet system the circulating water carries the heat, and the metal only conducts through a pipe wall a few millimetres thick. The poster also gave the alloy a conductivity of 401 W/m·K, which is pure copper’s; a typical aluminium–copper alloy is nearer 120.
• There were no numbers where they mattered. We had no room heat-loss estimate, heater output, installed cost per flat or running-cost comparison with radiators, although cost was one of our criteria.
• We presented an existing product as our own idea. Our script called ThermaSkirt a system we had innovated, but our own references show it’s DiscreteHeat’s product.

What I’d do differently

• Define each criterion, weight it properly and justify every score.
• Score insulation and heating together as packages.
• Estimate the heat loss of a typical Govan room and size the heater from manufacturer output data.
• Cost each flat, supply and installation, and compare annual running costs with radiators.
• Compare materials on conductivity, density, cost per kilogram and the embodied energy of recycled against primary aluminium.