← All guides

Case study — taking a 1920s house from EPC F to C

A real renovation diary — every efficiency project on a five-bedroom 1920s house, what each one cost, what actually worked, and what I’d do differently.

Most of the guides in this section explain what should work. This one is different: it’s the story of what I actually did to my own house over several years — every project, what it cost, what it changed, and the things I’d do differently with hindsight. One house is a sample size of one, so treat the numbers as illustrative rather than universal, but the ordering and the lessons travel well.

The house and the starting point

I bought the house in 2018: a brick-built 1926 property south of London with five bedrooms and just over 300m² across two floors, plus a converted loft room on a third floor and two remaining unheated loft spaces used for storage.

Its EPC rating at purchase was F — the second-lowest band there is. For context, the average UK home sits in band D, most new-builds achieve B, and a home can’t legally be let below band E. (You can look up any home’s certificate free at the EPC register on gov.uk — the EPC guide explains how the score works.)

The F rating showed up exactly where you’d expect — in the bills. The house was using around 53,000kWh of gas and 9,200kWh of electricity a year, with roughly a quarter of the electricity on a cheaper overnight rate via an old Economy 7 meter. Total running cost: about £2,900 a year at the prices of the day — the same consumption would cost around £5,300 today.

The to-do list came free with the certificate

The EPC report includes a ranked list of recommended improvements, and mine was a textbook case for an old house:

That list set my strategy: start with the small, high-impact projects and work up to the big ones. Here’s how it played out, roughly in order.

The cheapest band I ever bought: a second opinion

Before spending anything, I read the old certificate properly — and found that the assessor had assumed most of our walls had no cavity. They did. I paid a second assessor for a fresh report, and that correction alone moved the house from F to E, along with a revised to-do list.

This is the EPC guide’s point about evidence, learned the expensive way round: assessors can only credit what they can see or verify, and “as assumed” entries on a certificate are the tell. A survey fee bought a whole band before a single roll of insulation went in. Read your certificate before you read any quotes.

Quick wins: lighting and draughts

LED lighting was the easiest upgrade of all. The house was full of 50W halogen downlights, each of which swaps for a 4–7W LED equivalent in seconds. Replacing the lot — indoor and outdoor — was quick and cheap; today only the utility room and garage are still on old fluorescent tubes.

Draught-proofing came next: sealing strips around leaky doors and loft hatches. There will always be pressure differences across a house, and the cheapest way to lose heat is warm air escaping and being replaced by cold air from outside. For an old house, this is the best pound-for-pound DIY improvement there is — the insulation guide covers the full pecking order.

Loft insulation — £200 well spent

The lofts had only a few inches of tired vermiculite. I rolled out fresh 150mm mineral wool over the top across both loft spaces — about £200 in materials and a weekend of itchy work. Heat rises, so topping up thin loft insulation is almost always the most obvious project on cost-versus-benefit grounds, and I’d expect this alone to have made a real dent in the gas bills.

What I’d do differently: Due to the layout of my loft, I insulated most of the loft space with 150mm on top of the existing vermiculite and this left me short of the 270–300mm the insulation guide recommends. I might come back to this to add more insulation with floor boards on top for storage.

Heating controls

I replaced the old controls with a Hive smart thermostat. I’ll be honest: I don’t think it improved efficiency much by itself, but being able to monitor and control both heating and hot water from a phone is genuinely convenient, and good controls score EPC points too — see the heating controls guide.

What I’d do differently: Many modern smart thermostats can modulate the temperatures of the boiler, which means that it can run the boiler at a lower temperature when outside temperatures are not that cold (which is more efficient and requires less switching on/off of the boiler). My thermostat does not support this kind of modulation, and my boiler at the time did not support it either, but as I later did upgrade the boiler it would have been better to get a modulating thermostat. I might come back to this in the future.

Upgrading appliances

We had an old and particularly inefficient tumble dryer, which we replaced with a modern heat-pump model — far more efficient, since it recycles its own warm air rather than venting it. We changed most of the other appliances too, but the dryer had by far the largest impact.

If you want to know what an appliance actually costs you, a smart plug with energy monitoring will tell you for about £15 — worth doing before assuming which one is the villain.

The big one: boiler and hot water system — ~£8,000

The 20-year-old boiler for the heating system was on its last legs, fed by a gravity tank upstairs. We replaced the boiler with a modern condensing model, and at the same time replaced and relocated the hot water cylinder into one of the lofts, converting the system from gravity-fed to mains-pressure (an unvented cylinder).

At around £8,000 this was the most expensive project so far, and while it probably improved heating efficiency by perhaps 10%, I doubt it will ever pay for itself on fuel savings alone — it simply had to be done. The side benefits were real, though: much better hot water flow, and we could rip out two annoying booster pumps that had been propping up the showers.

What I’d do differently: this was the moment to seriously evaluate a heat pump for the whole house, and I completely missed it. If your boiler is dying, price both options before defaulting to like-for-like gas.

Doors and windows — ~£2,500

The original oak front door was a beautiful feature — and cold and leaky. A modern composite door (~£2,500) made a marked difference to temperatures downstairs. We were lucky on windows: the previous owner had already upgraded from single to double glazing. If they hadn’t, glazing would have been a major project on the list.

Getting off electric heating

Three rooms were sporadically heated with plug-in electric radiators or with electric underfloor heating. Since direct electric heat costs roughly four times as much per unit as gas, we had a plumber extend the central heating into those rooms with proper radiators — an unglamorous £2,000 job that took a real bite out of the electricity bill.

Cavity wall insulation — ~£2,500

We were lucky that a 1926 house had been built with a 50mm cavity at all. I hesitated over filling it — you can read plenty of reports of moisture and damp problems after botched installs, so a proper pre-survey and a guaranteed installer matter. In the end I went ahead, for around £2,500, and it made a clear difference to both comfort and the heating bills.

Floor insulation — comfort first, payback second

One downstairs room, a snug converted from a garage, sat on bare uninsulated concrete. When we renovated it with new wooden flooring we added 50mm of PIR insulation board underneath — and the transformation in winter was so striking that we decided to insulate every downstairs floor.

We combined the work with flooring we were replacing anyway: 100mm insulation boards between the joists under the new wooden floor in the living room, then the same treatment when the kitchen, dining room and hallway were renovated. Insulating every floor was definitely not cheap — I estimate the insulation work came to around £5,000 as part of the wider flooring jobs, though doing it DIY while a floor is up would cost far less — and I doubt it improved the house’s overall heat loss by more than 10%. But the quality of life downstairs in winter is on another level. This is a “do it while the floor is up anyway” project, not one to rip up good floors for.

Where that left the bills

By this point gas consumption had fallen 40% and electricity about 30%. The annual cost was around £3,800 — and only that high because energy prices had risen sharply in the meantime. The house was noticeably warmer while burning far less.

A fresh EPC assessment at this stage confirmed it on paper: from the corrected E, the house had climbed to band C — better than the UK average, and F to C overall, on fabric and heating work alone, before a single panel or battery was installed. Fabric done; time to work on the electricity side.

Battery and solar — ~£10,500

A home battery does two jobs: it charges overnight on a cheap tariff and runs the house through the day, and paired with solar it stores your own generation for when the sun isn’t shining. The economics get dramatically better if you have an EV or plug-in hybrid, because that unlocks the cheapest tariffs — the smart tariffs guide covers how these work.

My sequence: I got a PHEV, installed a smart-meter and EV charger and switched to Octopus’s Intelligent Go tariff (7p/kWh between 23:30 and 05:30). Then I installed a 20kWh battery with a 10kW inverter in the garage, charging overnight on the cheap rate — that alone slashed the electricity bill. Later I added twelve 645W solar panels (about 7.7kWp) into the same inverter, and started exporting the surplus.

Over the past 12 months my total electricity bill has been £45. That’s an annual saving of roughly £1,750 against a total install cost of £10,500 — a break-even of around six years, assuming prices and export rates hold. You can run your own numbers in the Solar & Battery ROI Simulator.

The honest verdict: if you have an EV or PHEV, the battery is by far the better half of the investment. Solar is great in summer but generates very little in a UK winter; the battery keeps working every night of the year, running the house on 7p electricity instead of ~28p. It’s also invisible, which the panels are not. See the solar and battery guide for the general economics.

What I’d do differently: Octopus has started launching V2H (Vehicle to house) and V2G (Vehicle to grid) solutions, where you charge your EV overnight and allow it to run the house during the day. At the moment this only works for a few car models, and as far as I know only with Octopus, but it is an interesting solution. You can for example get a second hand VW ID.4 with 77kWh battery for about 10-15k GBP, and this is much cheaper and larger capacity than most house batteries. The benefit of having a house battery is that it is always connected, but I would definitely have considered this option if I started this journey today.

The scorecard so far

Project Approx. cost Verdict
LED lighting Low £100s Trivial cost, instant saving — do first
Draught-proofing Under £100 DIY Best £-for-£ improvement in an old house
Loft insulation top-up ~£200 DIY The classic no-brainer — but lay the full 270mm if possible
Smart thermostat ~£200 Convenience more than savings
Boiler + unvented cylinder ~£8,000 Necessary, not an investment — consider a heat pump instead
Composite front door ~£2,500 Marked comfort improvement downstairs
Radiators replacing electric heaters ~£2,000 Cut the expensive fuel out
Cavity wall insulation ~£2,500 Clear win on comfort and bills
Floor insulation (whole ground floor) ~£5,000 (far less DIY) Comfort transformation; do during renovation
20kWh battery + 7.7kWp solar ~£10,500 ~£1,750/yr saved, ~6-year payback; battery is the star

The headline: annual running cost down from £2,900 to £1,950 (now essentially all gas) — achieved while my unit rates almost doubled, gas from 3.1p to 5.7p and electricity from 15p to 28p. Hold prices constant and the fall is far steeper; the scorecard below does that arithmetic properly. The house is warmer, cheaper and more comfortable than at any point since 1926, I’d wager.

What’s next: chipping away at the gas

That remaining £1,950 is almost all gas for heating, so the current phase is about displacing the boiler with heat pumps:

If it all works, the endgame is heat pumps heating (and cooling) the whole house from cheap overnight electricity, with the gas boiler reduced to a backup for the depths of winter.

Was it worth it?

Energy prices roughly doubled over these years, which makes the raw bills a poor scorecard — a saving can hide behind a rising unit rate. So the honest comparison is the last column: what 2018’s consumption would cost at today’s rates, against what I actually pay now.

Gas

Year Usage Cost paid Cost at today’s rate
2018 53,000kWh £1,700 £3,100
2026 32,000kWh £1,900 £1,900
Change −40% +12% −39%

Electricity

Year Usage Cost paid Cost at today’s rate
2018 9,200kWh £1,200 £2,200
2026 6,500kWh £45 £45
Change −30% −96% −98%

So the bill I actually pay went from £2,900 to £1,950 — but measured properly, on a like-for-like price basis, the 2018 house would cost £5,300 to run today against my £1,950. That is the real number: −63%, and it includes nightly PHEV charging that has taken a chunk out of the petrol bill too (approximately 980GBP based on charging car with 1.4MWh).

What else changed, beyond the money:

And once the A/C units are in, I expect gas heating to fall at least another 30%, summer heatwaves to become bearable, and the office extension — currently too hot in July and too cold in January — to be comfortable year-round.

The whole programme came to roughly £30,000 over several years, and it takes annual running costs from £5,300 to £1,950 at today’s prices — a break-even of about nine years. I can’t put a firm number on what it did to the house’s value, though I’d be surprised if it were less than the £30,000, and an EPC C house will be a great deal easier to sell than an EPC F one. On my own numbers, a no-brainer — with the caveat that a good half of this spending bought comfort rather than payback, and I’d have done most of it anyway.

One caveat to these calculations is that having a PHEV and a cheap overnight tariff is a core part of the economics around house battery and these savings. I have NOT included the cost of the PHEV in these calculations (it was probably marginally more expensive than the equivalent petrol/diesel car). At the same time, I have not directly included the petrol savings in the calculations – if I did that the real saving would be going from £5,300 to £970 or a saving of 82%.

The one-line summary

Start from the EPC’s own recommendation list, do the cheap fabric work first (lights, draughts, loft), take the big fabric opportunities when renovation opens them up (cavities, floors, doors), and once demand is down, a battery on a cheap EV tariff — with solar as the supporting act — is the single best-returning purchase in the whole journey.

Last reviewed: August 2026.