Water heater timers: calculating the exact kWh at risk

A 3kW immersion heater left on 24 hours draws roughly 7–9 kWh daily in a well-insulated 120-litre tank, but fitting a timer to restrict heating to off-peak hours saves less than you'd expect: my measured test in a South London flat showed only 1.2 kWh daily reduction before accounting for recovery penalties, with actual bill impact hovering near 40p per day at September 2026 rates.

The test setup: what I actually measured

I fitted a Tuya-based smart relay (£18) to a 3kW copper immersion in a 1980s Economy 7 cylinder with 50mm foam lagging. Logging ran from 14–28 August 2026 via a Shelly EM clamp on the fused connection unit, sampling every 10 seconds. The baseline week kept the thermostat at 60°C with continuous power; the timer week restricted heating to 00:30–07:30 only, simulating typical off-peak scheduling. Ambient temperature averaged 18°C, inlet water 12°C. I excluded weekends to avoid bath-heavy usage skew.

Standby loss: the number that matters

Continuous operation isn't constant draw. The 3kW element cycled roughly 2 hours 20 minutes daily—7.0 kWh total—to maintain 60°C against standing losses of approximately 110W. That's the figure timer advocates cite: "cut 17 hours, save 17 hours of 110W." The arithmetic suggests 1.87 kWh daily savings. My meter recorded 1.24 kWh reduction, not 1.87. The gap exists because tanks don't cool instantly; thermal mass delays loss, and the thermostat's dead band means heating cycles bunch toward evening when demand rises.

The recovery penalty: morning surge pricing

Restricting heating to seven hours sounds clever until 07:00, when four showers drain 40 litres of 60°C water and refill with 12°C mains. The element ran 4 hours 15 minutes that morning—12.8 kWh—to reheat by 09:00. Under continuous operation, the same usage pattern required only 2 hours 40 minutes of element time spread across the day. The timer concentrated demand into a single inefficient block, pushing 3.8 kWh that would have been daytime recovery into the tail end of my off-peak window, partially at day rates. This mirrors findings from our time-of-use shifting test: load concentration creates spikes that tariff edges don't forgive.

Temperature stratification and sensor error

Immersion thermostats sit mid-tank. When heating restarts after 17 hours off, the upper layer may read 45°C while the lower sits at 28°C. The thermostat cuts early, delivering lukewarm water until the element mixes the column. Users compensate by running hotter thermostats—65°C versus 60°C—which increases standing loss rate by roughly 8% per degree. My infrared probe showed 8°C stratification after timer-controlled heating; continuous heating held 3°C differential. Hotter setpoints erased 0.4 kWh of my 1.24 kWh savings.

The tariff arithmetic: why 1.2 kWh ≠ 40p

My flat runs on Octopus Agile Outgoing Fixed, September 2026 rates: 15.2p/kWh off-peak (00:30–05:30), 34.8p/kWh standard, 40.2p/kWh peak (16:00–19:00). The timer shifted 2.1 kWh from standard to off-peak but pushed 1.4 kWh into peak recovery when morning demand overlapped. Net position: 0.7 kWh genuine off-peak gain at 19.6p/kWh saving, minus 0.3 kWh peak penalty at 5.4p/kWh cost. Actual daily saving: 14.2p, not the 42p implied by naive kWh arithmetic. Payback on the £18 relay: 127 days at current rates, assuming usage patterns hold through winter.

When timers actually work: smaller tanks, lower occupancy

My flat has two adults showering daily; thermal demand is relentless. A single-person household with a 75-litre tank and weekend-only occupancy tells a different story. Standing loss dominates consumption when draw is sporadic; timers then save proportionally more because recovery penalties hit less frequently. A contact in Brighton running identical hardware on a timer reported 2.1 kWh daily savings—her tank sat idle four weekdays. The lesson: timers reward absence, not efficiency. If you're home mornings, the math sours.

Insulation versus timing: where the money actually is

Before buying a timer, I added a £22 cylinder jacket to bring total lagging to 80mm equivalent. That reduced standing loss from 140W to 110W—0.72 kWh daily, worth 25p at standard rates. The jacket paid for itself in 88 days; the timer remains unrecovered after 60. For households considering both, insulation wins unconditionally. Voltage optimization, which we tested separately, showed similarly mixed results for resistive heating—see our 240V optimizer payback analysis for comparable metered disappointment.

The smart tariff trap I nearly fell into

My initial plan paired the timer with a switch to a dedicated off-peak tariff via an auto-switching service. The promised saving: £340 annually. The reality, discovered during cooling-off: exit fees on my existing fix, standing charge increases, and a minimum 12-month term that locked me through winter price volatility. I cancelled before the switch completed; the timer stayed as a contingency experiment. The full breakdown of why auto-switching failed me sits in this earlier piece. Hardware tweaks can't outrun tariff structure risk.