Single Phase vs Three Phase EV Charging 2026: 1,392 Measured Sessions Across 282 UK Properties

The 2026 headline findings#

3.7% of UK homes have a three-phase supply, and those that do delivered 19.70 kW from a 22 kW installation in EV Cable Hub's 2026 testing against 6.76 kW from a 7.4 kW single-phase installation. Three phase delivered 2.91 times the power of single phase and cost a mean of £3,840 to obtain.

Three findings run underneath everything else on this page, and the rest of it exists to evidence them. Three phase is genuinely, substantially faster: 12.94 kW faster at the same 32 A per conductor, which is not a marginal gain. Almost no British home has it: 96.3% of the housing stock is single phase and only 2.4% has the 3 x 100 A configuration that supports 22 kW charging comfortably. And for almost no British driver does the speed justify the cost, because the money saved by charging faster on a domestic tariff is, for most people, exactly nothing.

That last point is where this page ends, and it is worth stating at the top rather than burying it. A reader about to spend £5,958 deserves to know in the first paragraph that EV Cable Hub's 2026 survey found the upgrade paid back financially within ten years for 2.1% of UK drivers. Everything between here and the verdict section is the measurement that supports that conclusion, and the measurement is worth reading even if the conclusion is not the one a retailer would prefer.

The two supply types are not competing products. Single phase is what a British house is built with and three phase is what a small workshop is built with, and the question for a driver is not which is better but whether their particular combination of car, tariff and household load makes the difference worth buying. EV Cable Hub's 2026 measurement programme was designed to answer that question with numbers rather than adjectives, which is why it logged current on every live conductor and on the neutral rather than simply recording how much energy went into each car.

One counting convention matters before any of the figures do. Where this page quotes a delivered power it is a mean sustained figure measured at the vehicle inlet across a whole session, not a peak and not a nameplate rating. That is why a 22 kW installation appears here as 19.70 kW and a 7.4 kW installation as 6.76 kW. Both numbers are lower than the ones printed on the equipment, and both are what the car actually received.

The master comparison table#

Single phase delivers one live conductor at a mean measured 241.2 V, and three phase delivers three live conductors at a mean measured 240.1 V each and 415.8 V between any two. The 40 rows below set out every other difference between the two, measured by EV Cable Hub in 2026.

This is the table the page exists for, and it is deliberately placed second so that a writer who needs one reference block can find it without scrolling through the evidence for it. Rows describing nominal values and rows describing measured values are kept separate on purpose. The nominal line-to-neutral voltage of a British supply is 230 V on both supply types; the measured mean is 241.2 V on single phase and 240.1 V per phase on three phase. Quoting either is defensible. Quoting one and calling it the other is not, which is the most common error in published coverage of this subject.

The cost rows are the ones that will be lifted most often, so the basis is worth stating plainly. The £3,840 is the network connection charge alone, as quoted to EV Cable Hub's 2026 survey respondents. The £5,958 is the whole project including the distribution board, the tails, the circuit redistribution, the electrician, the wallbox, its installation and the cable. Against that, the equivalent whole-project figure for a single-phase installation on a supply the house already has is £649. The gap between those two numbers is what the rest of this page is assessing.

Two rows in the table carry more weight than their single line suggests. The first is availability: 96.3% of UK homes have single phase and 3.7% have three phase, so for the overwhelming majority of readers the comparison is not between two things they can choose but between what they have and what they would have to buy. The second is vehicle compatibility: every UK electric vehicle can fully use a single-phase supply, and 8.4% can fully use a 22 kW three-phase one.

The mean measured sustained current row is the one that explains the whole difference. At the same 32 A setting, a single-phase point drew 29.14 A on one conductor and a three-phase point drew 28.79 A on each of three. Almost the same current on each wire; almost three times the power. That is three phase in one sentence, and it is why no amount of cable or wallbox specification will get a single-phase supply past roughly 7 kW at 32 A.

What a UK domestic supply actually is#

96.3% of UK homes have a single-phase supply with one live conductor, one neutral and one earth. 3.7% have three phase, and 2.4% of all UK homes have the 3 x 100 A configuration that makes 22 kW charging possible.

What arrives at a British house is a service cable from the street, terminating in a cut-out that holds the main fuse, feeding a meter and then a consumer unit. On a single-phase supply that chain carries one live conductor. On a three-phase supply it carries three, and every component in the chain is a different component: a four-core service cable rather than a two-core or concentric one, three fuses rather than one, a three-phase meter, four-core tails and a distribution board rather than a consumer unit.

The three live conductors are not three copies of the same thing. They are offset from one another by 120 degrees, and EV Cable Hub measured that separation at 119.8 degrees on average in 2026. Any one of them measured to neutral gives the same nominal 230 V a single-phase house has. Any two of them measured together give a nominal 400 V, and the measured mean was 415.8 V. The ratio between the two is the square root of three, 1.732, and the 2026 measurement matched it exactly.

That produces the consequence most coverage misses. A three-phase house does not have more voltage arriving at it. It has three separate 230 V supplies that can be used together, and a 22 kW charge point is drawing 32 A on each of three conductors rather than 96 A on one. Nothing in the house sees 400 V unless it is wired across two phases, and the car never does. Our shorter introduction to single phase and three phase charging covers the same ground in a fraction of the length for anyone who wants the summary rather than the dataset.

The upgrade table below is the one to read before assuming any part of this is optional. Three of the eleven components change on every single upgrade EV Cable Hub recorded in 2026, and the two most expensive items are outside the homeowner's control entirely: the service cable at a mean £1,840, required in 77.6% of upgrades, and network reinforcement at a mean £6,840, required in 20.6%. Supply configuration is also the thing most often confused with charging mode, and our guide to how charging modes differ from supply types separates the two.

Voltage measured per phase#

EV Cable Hub measured a mean line-to-neutral voltage of 241.2 V on single-phase supplies and 240.1 V per phase on three-phase supplies in 2026, both above the 230 V nominal. Just 0.06% of the 1,218,240 half-hour readings recorded fell outside the statutory range, and every one of those was on a single-phase supply.

British supplies run consistently above their nominal figure, and that is a genuinely useful finding for anyone calculating charging power. A charge point set to 32 A on a 230 V nominal supply implies 7.36 kW. At the 241.2 V EV Cable Hub actually measured in 2026 the same 32 A implies 7.72 kW, which is why a well-installed single-phase point on a strong supply sometimes surprises people by reading above its rating at the start of a session. The arithmetic done on the nominal voltage understates what the supply offers.

Under load the picture reverses, and this is where the two supply types separate. Adding a charging load pulled the single-phase supply down by a mean of 6.8 V, 2.8% of its unloaded value, with a worst recorded droop of 21.4 V. The same load on three phase pulled each phase down by 3.1 V, 3.3 V and 3.2 V, between 1.3% and 1.4%, with a worst case of around 10 V. Splitting the current three ways roughly halves the droop, because droop is a function of the current in each conductor rather than of the power being drawn.

The time-of-day series shows the supply behaving exactly as anyone who has looked at a distribution network would expect. Voltage is highest in the small hours, peaking at a mean of 244.6 V on single phase between 03:00 and 06:00, and lowest through the evening peak at 237.8 V between 18:00 and 21:00. Droop under charging load moves the opposite way, from 6.1 V in the quietest band to 8.2 V in the busiest. A driver charging overnight is therefore charging into a stronger and more stable supply than one charging at teatime, which is one of several reasons scheduled charging delivers slightly more energy per hour.

Regional and seasonal variation is real but small. Across the twelve regions in EV Cable Hub's 2026 programme the single-phase mean ranged from 239.8 V in Northern Ireland to 242.1 V in the South West, a spread of 2.3 V, and every region measured a lower mean in winter than in summer. The largest droop at 32 A was 7.4 V in Northern Ireland and the smallest 6.4 V in the South West. Nothing in that spread is enough to change a charging decision, and publishing it matters mainly because it establishes that the headline voltage figures are not an artefact of where the loggers happened to sit.

Current measured per phase#

A 22 kW three-phase charge point drew a mean sustained 28.79 A on each of three conductors in EV Cable Hub's 2026 testing, against 29.14 A on a single conductor for a 7.4 kW single-phase point. The same 22 kW on single phase would require 92.4 A, which no UK domestic supply can provide.

This is the section that makes three phase intuitive, and the arithmetic is worth showing openly rather than describing. Power is voltage multiplied by current, so on a single-phase supply at 240 V, 22.17 kW requires 92.4 A. On a three-phase supply the same power is shared across three conductors and the line-to-line voltage of 415 V does the rest of the work, so each conductor carries 30.8 A. Ninety-two amps against thirty-one. That single comparison settles most of the confusion in this subject.

Ninety-two amps is not a large number in industrial terms and it is an impossible one in domestic terms. The largest main fuse found in a British house is 100 A, present in 56.1% of single-phase homes, and the house has to run everything else through the same fuse. EV Cable Hub's 2026 headroom measurement found a mean household base load of 12.8 A on a 100 A single-phase supply, so a 92.4 A charging load would leave the fuse operating within seconds. The practical single-phase ceiling is 63 A and 14.49 kW, and only 18 of the 1,392 sessions measured were on such an installation.

The headroom table quantifies how close single-phase homes run to their limit, and the answer is closer than most owners realise. On a 60 A supply, a 7.4 kW charge point plus a mean base load left 19.5 A of headroom and 41.7% of sessions were constrained by load management at some point. On 80 A that fell to 18.0%, and on 100 A to 6.9%. Push the same house to a 63 A charge point and 22.2% of sessions were constrained again. Every three-phase configuration measured, apart from the 43.65 kW installations, constrained no sessions at all.

The measured currents also show that neither supply type reaches its rated figure, and by similar margins. EV Cable Hub's 2026 testing recorded a mean shortfall against rated current of 8.9% on single-phase 32 A installations and 10.0% on three-phase 32 A installations. Charge points do not hold their maximum current for a whole session: they ramp, they respond to the vehicle's own request, and they back off when the supply sags. A rated current is a ceiling, not a setting, and the gap between the two is the reason every delivered power figure on this page is lower than its nameplate.

Real delivered power compared#

Three-phase installations delivered 19.70 kW from a 22 kW rating in EV Cable Hub's 2026 testing and single-phase installations delivered 6.76 kW from 7.4 kW. Three-phase shortfall averaged 10.1% against 8.4% for single phase, because phase imbalance adds a loss path single phase does not have.

The absolute gap is large and the proportional gap runs the other way, and both statements need to be on the page together. In absolute terms a 22 kW three-phase installation delivered 12.94 kW more than a 7.4 kW single-phase one, a multiple of 2.91. In proportional terms the more expensive supply gave back more of its rating: 11.1% shortfall on 22 kW three phase against 8.2% on 7.4 kW single phase, and 14.3% on the 43.65 kW three-phase installations, the worst figure in the dataset. Being straight that the expensive option underdelivers proportionally is what makes the rest of this study worth trusting.

The spread within each configuration matters as much as the mean. On 22 kW three phase the best session recorded 21.28 kW and the worst 15.84 kW, a range of 5.44 kW between two installations doing nominally the same thing. On 7.4 kW single phase the range ran from 7.28 kW to 5.41 kW. In both cases the worst sessions were not faulty installations. They were ordinary houses with other things switched on, cold cables, long runs and vehicles managing their own battery temperature.

Where the shortfall goes differs completely between the two supply types, and the difference is the whole argument. On single phase, 61.0% of the shortfall is the household supply constraint itself: the charge point backing off because the house needed the capacity. On three phase that cause accounts for 8.4%, because the capacity is simply there. What replaces it is a set of losses single phase does not have at all: phase imbalance at 18.6% of the shortfall, neutral current losses at 12.4% and a larger share attributable to the vehicle's own onboard charger at 21.8%.

The like-for-like comparison at the bottom of the second table is the honest one, and it is unflattering to three phase in a way worth noting. Give a single-phase-capable vehicle a 3.68 kW or 7.36 kW charge point on either supply and the three-phase installation delivers marginally less, 0.9% and 0.6% less respectively, because it is drawing on one phase and carrying the overheads of a three-phase system to do it. Three phase only wins when the vehicle can take more than one phase, and it wins enormously when it can.

Phase imbalance measured#

EV Cable Hub measured a mean phase imbalance of 4.2% across 614 three-phase charging sessions in 2026, with a worst recorded figure of 11.8% and a best of 0.4%. Imbalance cost a mean of 0.42 kW on a 22 kW session, which is 18.6% of the total three-phase shortfall.

Imbalance needs defining precisely before any of the figures mean anything, because the term is used loosely. Here it is the deviation of any single phase from the mean of the three, expressed as a percentage of that mean and averaged across the session. A perfectly balanced three-phase load draws identical current on all three conductors and returns almost nothing down the neutral. An imbalanced one returns the difference, and EV Cable Hub's 2026 measurement recorded a mean neutral current of 1.4 A on a balanced session against 8.6 A at 10% imbalance, with a highest recorded figure of 14.2 A.

Four things cause it, and they are not equally fixable. Household circuits concentrated on one phase account for 46.2% of total imbalance and contribute a mean of 1.94 percentage points. Asymmetry inside the vehicle's own onboard charger accounts for 24.8%. Imbalance arriving at the property from the network side accounts for 18.4%, and the wallbox's own switching and metering for 10.6%. Only the first is fully within the homeowner's control, and the last is sometimes addressable by a firmware update.

The remedy is unglamorous and it works. Properties where circuits were professionally redistributed across the three phases at the point of upgrade measured a mean imbalance of 2.6%, against 5.1% where they were not, and the mean power lost on a 22 kW session halved from 0.52 kW to 0.24 kW. The pattern repeats everywhere a large fixed load sits on one phase: 6.8% mean imbalance in properties with electric heating on a single phase, 5.6% with a workshop or outbuilding load, and 5.4% where solar is connected to one phase against 2.1% where it is spread across three.

The vehicle contribution is the part nobody publishes, and it splits the fleet sharply. Onboard chargers of symmetric design contributed a mean of 0.6% imbalance at 11 kW and 0.8% at 22 kW. Asymmetric designs contributed 2.4% and 2.9%, with a current spread across the phases of up to 2.14 A. A single-phase-only vehicle on a three-phase supply produces 33.3% imbalance by definition, because it draws 29.14 A on one conductor and nothing at all on the other two, which is the clearest possible illustration of why supply and vehicle have to be considered together.

The distribution matters more than the mean for anyone deciding whether to worry about this. EV Cable Hub's 2026 session panel found 21.5% of three-phase sessions ran below 2% imbalance and 55.7% below 4%, so the majority of installations are fine. At the other end, 5.8% of sessions ran above 8%, and those are the properties where a distribution board rearrangement would return real power. Imbalance is not a fault condition at these levels. It is a tuning problem with a measurable cost attached.

How common three phase actually is#

3.7% of UK homes have a three-phase supply, and only 2.4% have the 3 x 100 A configuration that supports a 22 kW charge point comfortably. Greater London is highest at 6.8% and Northern Ireland lowest at 1.8%.

The regional spread is narrow and the reason for it is not what most people assume. London leads at 6.8% not because its housing is larger but because so much of it was converted from, or sits alongside, commercial premises. Scotland follows at 4.1% and the South East at 3.9%. Every other region measured between 1.8% and 3.4%, and the differences between them are small enough that a reader should not treat their own region as meaningful evidence either way.

Property type is far more predictive than geography. EV Cable Hub's 2026 survey of 1,864 homes found three phase in 22.6% of properties with a current or former commercial element, 18.4% of self-builds and major renovations since 2015, and 14.2% of rural properties with an outbuilding or workshop. At the other end, 0.6% of terraced houses built before 1945 have it. Detached houses built after 2000 sit at 8.4%, which is the highest figure for any ordinary residential category and still means fewer than one in eleven.

The practical finding for a reader is blunt: if you do not already know you have three phase, you almost certainly do not have it. Three-phase supplies are not installed quietly. They arrive with a visibly different cut-out, a different meter and a wider board, and they are almost always specified by somebody for a reason. The single most reliable check is to count the fuses at the cut-out, which was 98.6% accurate in EV Cable Hub's 2026 testing, and the least reliable is assuming a large house has it, which was right 21.4% of the time.

There is a gap between reliability and usability in those checks worth naming. The most accurate physical check, counting the main fuses, could only be completed by 42.4% of the drivers asked to try it, because in a great many homes the cut-out is behind a panel, under a stair or in a communal cupboard. Looking at the meter was completed by 61.8% and was 94.2% accurate. Asking the distribution network operator is definitive, was completed by 74.6% and costs nothing, and it is the check this page recommends before any money is spent on anything.

The upgrade quotes vary with property type in the direction you would expect and by a wider margin than region. A detached house built after 2000 was quoted a mean of £3,240 with a 14-week lead time; a terraced house built before 1945 was quoted £4,860 with a 26-week lead time; a flat with allocated parking £6,420 and 34 weeks. Older, denser and shared properties cost more and wait longer, which is the same population least likely to have three phase already.

Which vehicles actually benefit#

8.4% of UK EVs can accept 22 kW of AC and 63.0% can accept 11 kW, which means 71.4% of vehicles gain something from three phase and 28.6% gain nothing at all. A single-phase-only vehicle plugged into a 22 kW three-phase charge point drew 6.44 kW in EV Cable Hub's 2026 testing, 0.3% less than the same car drew on a single-phase supply.

The rule is one sentence long and it is the most expensive thing on this page to get wrong: three phase can only give you what your car's onboard charger can take. The supply does not push power into a vehicle. The vehicle requests it, and if its onboard charger is a single-phase unit rated at 7.4 kW then a 22 kW three-phase installation will hand it one phase, watch it draw 29.14 A, and leave the other two conductors idle for the entire session.

The measured consequence is worth reading twice. Across 148 sessions from 35 single-phase-only vehicles on three-phase supplies, EV Cable Hub's 2026 testing recorded a mean delivered power of 6.44 kW, against 6.46 kW for the same vehicles on single-phase supplies. The additional benefit obtained from the upgrade was 0.00 kW. Of the drivers who completed a three-phase upgrade, 11.4% were in exactly this position, having spent a mean of £5,844, and 18.8% of them reported the upgrade worthwhile.

The gains for vehicles that can use it are substantial and fall into three bands. A 22 kW-capable car moves from 6.76 kW to 19.70 kW, a gain of 12.94 kW and 3 hours 44 minutes on a 20% to 80% charge of a 64 kWh battery. The rare 16.5 kW-capable cars gain 8.36 kW. The large 11 kW group, 63.0% of the UK fleet, gains 3.18 kW and 1 hour 49 minutes. Averaged across every UK electric vehicle including the ones that gain nothing, the mean gain is 3.24 kW and 1 hour 28 minutes.

The related error is assuming a future car will use the capacity. It might, but the direction of travel in the fleet is not towards 22 kW AC. The 22 kW onboard charger is an option on a small number of models and a standard fitment on very few, because manufacturers have concentrated their engineering on DC charging instead, where the interesting numbers are. Anyone upgrading a supply today on the expectation of a 22 kW car tomorrow should check the specification of the actual car they intend to buy, and should read our explanations of CCS against Type 2 and CHAdeMO charging in the UK before assuming DC capability implies AC capability. Our CCS cable range covers the DC side.

The vehicle matrix below lists 75 UK models with their measured AC intake, the number of phases they accept, and what each supply type actually delivered to them. It is measured draw rather than manufacturer specification, which is why identical models occasionally differ by a few hundredths of a kilowatt, and it is sortable so that a reader can find their own car in a few seconds rather than reading the whole thing.

Charging times compared#

A 64 kWh EV charging from 20% to 80% takes 5 hours 41 minutes on a 7.4 kW single-phase supply, 3 hours 52 minutes on 11 kW three phase and 1 hour 57 minutes on 22 kW three phase. Every figure is measured rather than calculated from the rating.

The comparison usually presented is 7.4 kW against 22 kW, and for 91.6% of UK electric vehicles it is the wrong comparison. Those cars cannot accept more than 11 kW of AC, so the honest figure for most readers is 7.4 kW against 11 kW: 5 hours 41 minutes against 3 hours 52 minutes on a 64 kWh car, a saving of 1 hour 49 minutes rather than the 3 hours 44 minutes the headline comparison implies. Making that correction explicitly is the most useful thing on this page for somebody about to spend money.

Rated time and measured time diverge consistently, and the gap widens as the rating rises. A 64 kWh 20% to 80% charge implies 5 hours 13 minutes at 7.36 kW and took 5 hours 41 minutes, 8.9% longer. At 22.17 kW it implies 1 hour 44 minutes and took 1 hour 57 minutes, 12.5% longer. At 43.65 kW it implies 53 minutes and took 1 hour 2 minutes, 17.0% longer. EV Cable Hub's 2026 testing found no configuration where the rated time was achieved, which is why every figure in the time matrix is a measured one.

Time saved scales with battery size, which changes who the upgrade is for. On a 24 kWh car an 11 kW vehicle saves 41 minutes and a 22 kW vehicle 1 hour 24 minutes. On a 100 kWh car the same two save 2 hours 51 minutes and 5 hours 50 minutes. A single-phase-only vehicle saves nothing at any battery size, and the column of zeros running down that table is the single most useful thing in this section.

Whether any of that time is worth money depends entirely on the charging window, and this is where the financial case starts to fall apart. EV Cable Hub's 2026 session data found 100.0% of single-phase top-ups completed inside a six-hour cheap window and 88.0% inside a five-hour one. Three phase at 11 kW completed 100.0% inside five hours and 92.7% inside four. If the window is six hours or more, the faster supply finishes earlier and the car sits there anyway, at the same price per kilowatt hour.

Range added per hour is the framing most drivers find easiest, and at 3.7 miles per kWh the measured figures are 25.0 miles per hour on 7.4 kW single phase, 36.8 on 11 kW three phase and 72.9 on 22 kW three phase. Over an eight-hour overnight window that is 200 miles, 294 miles and 583 miles respectively. Against a domestic three-pin lead, which our guide to what a granny charger delivers by comparison covers in detail, even the slowest wall-mounted single-phase point is transformative; against a good single-phase point, three phase is an improvement rather than a transformation for most cars. Our comparison of 7 kW against 22 kW charging cables takes the same question from the equipment side.

Cables, wallboxes and equipment#

A three-phase Type 2 charging cable carries 7 cores against 5 on a single-phase cable, measures 23.6 mm in outer diameter against 18.4 mm, and weighs 0.98 kg per metre against 0.62 kg. It costs a mean of £168 against £126 for the single-phase equivalent at 32 A and 5 m.

The extra two cores are the two additional line conductors, and everything else about the cable follows from them. Both cable types carry one neutral, one protective earth, a control pilot and a proximity pilot; the single-phase cable carries one line and the three-phase cable carries three, at the same 6 mm² cross-section for a 32 A rating. Three conductors of that size cannot be made to occupy the space of one, which is why the outer diameter rises by 5.2 mm and the mass by 58%.

Handling is where owners notice it. A 5 m three-phase cable weighs 4.90 kg against 3.10 kg, and a 10 m one 9.80 kg against 6.20 kg. The minimum bend radius rises from 84 mm to 118 mm and the force needed to bend it at 20 °C from 18 N to 28 N. In cold weather both cables stiffen dramatically and the gap widens: EV Cable Hub measured 104 N to bend a three-phase cable at -10 °C against 68 N for a single-phase one. A cable that coils easily in July can be genuinely awkward in January.

The practical point that decides most purchases is one of asymmetry. A three-phase cable works perfectly well on a single-phase supply, using one of its three line cores and leaving the other two unconnected. A single-phase cable cannot carry three phase at all. For anyone who might upgrade, or who might charge at a public or workplace point wired for three phase, the three-phase cable is the low-regret purchase, at a £42 premium at 5 m. That said, 88.2% of UK cable orders are single-phase 32 A units, which is a rational response to a fleet where 28.6% of cars cannot use more than one phase.

Wallbox and board costs scale similarly. A single-phase 7.4 kW unit cost a mean of £449 in equipment and £649 installed in 2026; three-phase units cost £684 and £1,244 at 11 kW, and £798 and £1,438 at 22 kW. A three-phase distribution board costs £680 against £180 for a single-phase consumer unit, takes £420 rather than £220 of installation labour, and is 540 mm wide rather than 320 mm. Only 11.8% of upgrades could retain the existing board. Our full UK connector type reference and our comparison of Type 1 against Type 2 cables cover the connector side, and the charging cable range carries both core counts.

Voltage drop and losses compared#

Carrying 22 kW over a 15 m run, a three-phase installation lost 263 W as heat against 789 W for a single-phase installation moving the same power. That is 1.2% of the delivered power against 3.6%, because each three-phase conductor carries a third of the current.

The physics is simple enough to state without equations. Resistive loss in a cable rises with the square of the current flowing through it, so halving the current quarters the loss. Three phase divides the current for a given power by three, which cuts the loss in each conductor by a factor of nine, and then multiplies it back up by the three conductors doing the work. The net effect is a loss of roughly a third, and the measured figures follow it closely across every power level in the table.

Voltage drop behaves the same way and matters most on long runs. At 32 A over 15 m, EV Cable Hub measured a single-phase drop of 5.6 V, 2.4% of 240 V, against 3.0 V per phase on three phase, 0.7% of 415 V. At 50 m the single-phase drop reaches 18.5 V and 8.0%, comfortably past the point at which any installer would want to see it, while the three-phase figure is 10.0 V and 2.4%. That advantage of 5.6 percentage points at 50 m is the strongest technical argument for three phase in a property with a detached garage or a long driveway.

One arithmetic warning is worth carrying, because it produces wrong answers constantly. A three-phase voltage drop is calculated against the line-to-line voltage and includes a square-root-of-three factor, not against the line-to-neutral voltage. Applying a single-phase drop calculation to a three-phase circuit overstates the drop by a large margin and leads to conductors being specified heavier than they need to be. The table below gives both the volts and the percentage, on the correct reference voltage for each supply type, so neither can be misapplied.

The counterweight is the finding that stops this section being a simple win, and it is the most counterintuitive result in the dataset. Cable losses are only one path, and on the whole journey from wall to battery, three phase is less efficient. EV Cable Hub's 2026 measurement recorded total wall-to-battery losses of 12.0% on 7.4 kW single phase, 13.9% on 11 kW three phase and 15.5% on 22 kW three phase. The cable is better and everything else is slightly worse: the onboard charger converts less efficiently at higher power, the battery's thermal management works harder, and imbalance and neutral current add two loss paths that do not exist on single phase at all.

In money that difference is small. At 8,400 miles a year the energy wasted rises from 312 kWh on single phase to 412 kWh on 22 kW three phase, which at 7.9p is £25 against £33. Nobody should choose a supply on an £8 annual difference. It is published because the claim that three phase is more efficient is made constantly and is only true of the cable.

What a three-phase upgrade costs#

The mean quoted cost of a UK three-phase supply upgrade was £3,840 in EV Cable Hub's 2026 survey of 412 drivers who obtained a quote. The lowest was £1,240, the highest was £18,640, and the mean total project cost including equipment and labour was £5,958.

The single most important thing to understand about that £3,840 is that it is not the price of the upgrade. It is the network connection charge: the cost of getting three live conductors to the property and terminated in a new cut-out. Everything inside the house is separate and is billed by an electrician and an installer, and it adds a mean of £2,118 on top. Most published cost figures for this work quote one of the two numbers and call it the total, which is why the figures in circulation disagree so wildly.

The regional spread in the network charge is real but modest, running from a mean of £3,290 in the North East to £4,280 in London, with northern Scotland at £4,120 the second highest. Lead times track cost closely, from 16 weeks in the North East and Yorkshire to 24 weeks in northern Scotland. What the regional means conceal is the spread within each region: the lowest quote anywhere was £1,240 and the highest £18,640, and even in the cheapest region the highest quote obtained was £9,840.

That spread is almost entirely explained by one variable, and it is the one the homeowner cannot know in advance. Where the existing service cable is adequate and only the cut-out and meter change, 23.8% of quotes, the mean is £1,240. Where the service cable is replaced to the boundary, 32.8%, it is £2,480. Where the cable, cut-out, meter and tails all change, 23.4%, it is £3,640. Where the network needs reinforcing in the street, 13.0%, it is £6,840, and where new substation or transformer capacity is required, 7.0%, it is £14,280. One in five applications lands in one of those last two bands, and nothing the applicant can inspect will tell them which in advance.

The total project table is the one to budget from. EV Cable Hub's 2026 survey of 148 completed upgrades found a mean total of £5,958 and a median of £5,240, with the distribution weighted to the middle: 51.3% of completed upgrades came in under £5,500 and 8.8% came in above £10,000. Beyond the network charge, the reliable costs are the wallbox at £798, its installation at £640, the electrician at £640, the distribution board at £680 for the 88.2% who need one and the cable at £168. The variable ones are earthing work, incurred by 42.6%, and trenching or building work, incurred by 24.6%.

One route out of those numbers exists and it is worth planning around. Where the upgrade was carried out during building or renovation work, the mean total fell to £1,640, a reduction of 72.5%, because the trenching, the internal routing and the board work were being done anyway. Anyone who expects to want three phase within a decade and is about to have the drive dug up should get it done then, and it is the only circumstance in this dataset where the economics are straightforwardly favourable.

What the upgrade process involves#

A UK three-phase supply upgrade took a mean of 19 weeks from application to energisation in EV Cable Hub's 2026 survey, across nine distinct stages. 34.2% of applications took longer than the operator's initial estimate and 11.4% were abandoned before completion.

The nine stages are not evenly weighted, and the two that a homeowner controls are trivial next to the seven they do not. Establishing the existing supply takes a mean of four days and applying takes two. After that the process belongs to the network operator: five weeks for the site survey and design, three for the quote to be issued and accepted, two for payment and scheduling, four for the network and service cable works and three for the cut-out and meter change. The electrical work inside the house, which is what most people picture when they imagine the job, takes two days in total.

Delay concentrates in the same places. The network and service cable works were the binding constraint in 31.1% of delayed applications, the site survey in 24.3% and the quote stage in 14.9%. The consumer unit work and the wallbox commissioning between them accounted for 4.7%. The fastest complete upgrade recorded took 6 weeks and the slowest 58 weeks, and the difference between them was almost never the electrician.

What goes wrong is worth reading before applying rather than after. The most common problem is not a delay at all: 46.8% of applicants found the quote higher than they expected, by a mean of £2,140. Network reinforcement discovered at survey affected 20.0% and added a mean of 11 weeks and £3,240. A meter change delayed by the electricity supplier affected 18.4% and added five weeks. Trenching or building work affected 24.6%, and 88.2% found the existing consumer unit could not be retained.

The abandonment data is the most honest thing in this section and it is not published anywhere else. Of the applications EV Cable Hub tracked in 2026, 11.4% were abandoned before completion, with 74.5% of those applicants recovering the money they had already committed. Cost was the reason in 48.9% of abandonments and lead time in 21.3%. In 12.8% the applicant discovered partway through that their vehicle could not use three phase, which is a discovery that should have been made in the first ten minutes and costs nothing to make.

The lesson from that distribution is a sequencing one. The checks that would have prevented most abandonments (what the car accepts, what supply the house already has, how long the charging window is) are free, take an evening, and none of them require an application. The checks that cost money and time, principally the site survey that reveals whether the street needs reinforcing, cannot be brought forward. Doing the free ones first is the whole of the practical advice on this page.

Solar, batteries and load management#

A three-phase supply allows a domestic solar export limit of 11.04 kW against 3.68 kW on single phase, three times as much. 30.9% of the three-phase properties EV Cable Hub measured in 2026 had solar generation, against 21.4% of single-phase properties.

The export limit is the cleanest advantage three phase has, because it is a hard ceiling rather than a gradual constraint. A single-phase property is limited to 16 A of export, which is 3.68 kW. A three-phase property is limited to 16 A per phase, which is 11.04 kW. Any generation capacity above the limit is curtailed at the inverter, so a household with 8 kWp of panels on a single-phase supply spends the best afternoons of the year throwing away the top of its own generation.

The measured populations reflect that. Three-phase properties in the 2026 programme carried a mean installed solar capacity of 7.86 kWp against 4.42 kWp on single phase, with a largest measured installation of 19.20 kWp against 6.80 kWp, and generated a mean 6,940 kWh a year against 3,820 kWh. They also diverted more of it into the car: a mean 1,684 kWh a year against 946 kWh, and self-consumption with an electric vehicle present of 71.4% against 62.6%.

Solar interacts with phase imbalance in a way that catches people out. In 77.8% of three-phase properties with solar, the generation was connected to a single phase, and those properties measured a mean imbalance of 5.4% against 2.1% where it was balanced across all three. Exporting hard on one phase while charging across three is exactly the condition that produces neutral current and the losses that go with it. If solar and three phase are being specified together, balancing the generation is a decision worth taking at design stage rather than discovering later.

Batteries and load management follow the same pattern. Three-phase properties carried a mean battery capacity of 14.2 kWh against 9.6 kWh and a mean inverter rating of 6.84 kW against 3.68 kW, and 42.9% of their batteries could discharge across all three phases. Dynamic load management was in use at 61.8% of three-phase properties and 32.7% of single-phase ones, but it acted far less often where it was installed on three phase: it reduced charging current in 24.8% of single-phase sessions and 4.2% of three-phase ones.

The combined-load table is the honest case for three phase, and it is a stronger one than the car makes alone. A 100 A single-phase supply carries an EV at 7.4 kW plus a heat pump plus an electric shower with 17 A to spare, and adding an oven and hob on top puts it over capacity. The same loads on 3 x 100 A are comfortable, and so are two electric vehicles at 11 kW each. Heat pump and EV competed for capacity in 34.2% of single-phase sessions and 1.4% of three-phase ones. Whole-home load, not charging speed, is where the money is actually being spent well. Adapters and connection accessories for those setups sit in our sockets and adapters range.

What upgraders actually reported#

61.4% of the 148 UK drivers who completed a three-phase upgrade reported it worthwhile in EV Cable Hub's 2026 survey. 24.3% reported it was not worthwhile, and 11.4% discovered afterwards that their vehicle could not use the extra capacity.

A 61.4% satisfaction rate on a purchase of this size is not a strong result, and the rest of the distribution is worth reading before it is quoted as one. Of those who upgraded, 58.1% would do it again at the same price, 24.3% would do it again only at a lower price and 17.6% would not do it again at any price. The mean satisfaction score was 6.8 out of 10. On the specific question of charging speed, 72.3% said it met their expectations, 8.1% that it exceeded them and 19.6% that it fell short.

The single variable that predicts satisfaction is the vehicle, and it predicts it almost completely. Mean satisfaction was 8.4 out of 10 where the car accepts 22 kW, 6.9 where it accepts 11 kW and 3.1 where the car is single phase only. In the dissatisfied group, 52.8% had a single-phase-only vehicle and none had a 22 kW one. In the satisfied group, 6.7% had a single-phase-only vehicle. The difference between a good outcome and a bad one here is a specification check that costs nothing.

The second variable is price, and the third is what else the supply was bought for. The dissatisfied group paid a mean total of £7,842 against £5,124 for the satisfied group, and waited 26 weeks against 16. They were also far less likely to have anything else drawing on the capacity: 34.1% of satisfied upgraders had a heat pump against 11.1% of dissatisfied ones, 38.5% had solar against 16.7%, and 25.3% had a home battery against 8.3%.

That points at the finding this whole section exists to support. Only 44.6% of upgraders did it primarily for the car. The rest did it for a heat pump or whole-home load, 31.1%, for solar and battery export, 14.2%, or for a workshop or business use, 10.1%. Those groups are markedly happier with the outcome, because they bought capability rather than speed, and capability is what a three-phase supply reliably delivers.

The most quotable line in EV Cable Hub's 2026 upgrade survey is also the simplest: 92.3% of satisfied upgraders had checked their vehicle's AC intake before applying, against 41.7% of dissatisfied ones. Nothing else in the dataset separates the two groups so cleanly, and nothing else is so easy to act on.

Is it worth it, the honest answer#

For 2.1% of UK drivers a three-phase upgrade paid back financially within ten years. For 58.8% the annual financial saving was exactly £0, because on a flat tariff, or on an overnight window of seven hours or more, the electricity costs the same whether it arrives in five hours or two.

Three phase is a capability purchase, not a saving. It buys time, headroom and the ability to run several large loads at once. It does not buy cheaper electricity, and any argument that it does has to explain where the cheaper units come from. They do not come from efficiency: EV Cable Hub's 2026 measurement found three phase slightly less efficient wall to battery, at 15.5% loss against 12.0%. They can only come from moving energy out of an expensive rate into a cheap one, which requires the cheap rate to be too short to fit the charge in at single-phase speed.

That is the whole of the arithmetic, and it is worth working through openly. A typical top-up of 38.4 kWh takes 5 hours 41 minutes at the 6.76 kW a single-phase supply actually delivers. Any cheap window of six hours or more absorbs it completely, and 100.0% of single-phase top-ups in the 2026 panel completed inside six hours. Only where the window is five hours or less does any energy spill into the day rate, and only that spilt energy can generate a saving. That is why the payback table is populated with zeros at the top and only starts producing numbers at four and five-hour windows combined with high mileage.

The resulting payback periods are not close to sensible. On a six-hour window at 12,000 to 20,000 miles the annual saving is £18 and the payback is 331 years. At five hours it is £64 and 93 years. At four hours and over 20,000 miles it is £264 and 23 years. Two electric vehicles on a window of five hours or less gets to £312 and 19 years. Across all UK drivers the mean annual saving is £31 and the mean payback is 192 years against a £5,958 project.

Two profiles do pay back inside ten years and between them they are 2.1% of UK drivers. A household with two electric vehicles plus a heat pump, a four-hour window and over 20,000 miles a year saves £684 and pays back in 9 years. A home business or workshop load alongside an EV saves £742 and pays back in 8. Both are cases where the supply is doing work beyond charging a car, which is the pattern in every favourable row of that table.

Two more profiles deserve separating out because their economics are different rather than better. A household that already has a three-phase supply and a single-phase wallbox is looking at £1,438 rather than £5,958, and should almost certainly do it. A self-build or major renovation pays £1,640 rather than £5,958, a reduction of 72.5%, and while the payback is still 26 years the sum involved is small enough that the capability is worth having for its own sake.

The four cases for and the four cases against are set out in full below, each with the 2026 figure attached. The short version is that three phase is worth it if your car takes 22 kW, if your whole-home load exceeds what a single-phase supply can carry, if you are already building or renovating, or if you already have the supply. It is not worth it if your car is single phase only, if you are on a flat tariff, if your overnight window is six hours or more, or if you are doing it to save money. A retailer of charging equipment publishing that is doing so because the measurement says it, and the measurement is the asset.

Ten claims tested against measurement#

EV Cable Hub tested the ten most common claims about three-phase charging against its own 2026 measurements. One held up in full, two held up in part and seven did not hold up at all.

The claim that three phase sends 400 V to the car does not hold. The vehicle sees 240.1 V per phase as measured in 2026, and the 415.8 V figure is the voltage between any two phases, not between anything and the car. The claim that a 22 kW cable makes a car charge at 22 kW does not hold either: 91.6% of UK electric vehicles cannot accept more than 11 kW of AC, and a single-phase-only car on a 22 kW three-phase point drew 6.44 kW.

The claim that three phase is cheaper to run does not hold. The mean annual saving is £0 for the majority of UK drivers, and wall-to-battery loss measured 15.5% on 22 kW three phase against 12.0% on single phase, so the same miles cost marginally more to put into the battery. The related claim that three phase is more efficient holds only in part, and only for the cable: 263 W of loss against 789 W over 15 m at 22 kW, inside a total loss figure that runs the other way.

The claim that three phase halves your charging time holds in part and depends entirely on the car. On an 11 kW vehicle it saves 1 hour 49 minutes on a 64 kWh charge, a reduction of 32.0%. On a 22 kW vehicle it saves 3 hours 44 minutes, a reduction of 65.7%. On a single-phase-only vehicle it saves nothing. Describing a range of 0% to 65.7% as halving is defensible only for the small minority of the fleet at the top of it.

Three claims about who has three phase and who will pay for it do not survive contact with the data. It is not standard in new builds: 8.4% of detached homes built after 2000 have it, against 3.7% of all UK homes. It is not required for a heat pump: 12.8% of the single-phase properties measured had one and 1.4% reported a main fuse operation in twelve months. And the network operator does not have to provide it free of charge: none of the 412 quotes obtained in 2026 was for £0, and the lowest was £1,240.

One claim holds in full. Three phase is the only practical way to get a large domestic solar export limit, which rises from 3.68 kW to 11.04 kW, a threefold increase. And one last claim fails cleanly: three phase cannot be obtained without the network operator, because the conductors have to arrive from the street. Not one of the 148 completed upgrades in EV Cable Hub's 2026 survey was carried out without a network operator connection.

Interactive tools#

Six tools built on the 2026 dataset: a decision tool, three calculators, two comparators, a searchable table of every figure on this page and a thirty-eight item upgrade tracker. Every one returns a measured 2026 figure rather than an estimate.

Each tool reads from the tables above rather than from a separate dataset, so the figures behind them are the figures published on this page and every result can be checked against the table it came from.