The Great Britain voltage cascade from 400 kilovolts to 230 volts, and the per-unit maths behind it in May 2026

Where voltage management stands in May 2026

These notes stay short on news and long on mechanics, but a few changes in the last twelve months are worth setting down before the maths. The connections queue reform brought Gate 2 outcomes in April 2026 (283 gigawatts of generation and storage and 99 gigawatts of demand progressed to firm offers), which means more new connections at 132 kilovolts and 33 kilovolts than the network has seen in any twelve-month window since the 1965 super-grid build.¹⁰ ENA Engineering Recommendations G98 and G99 are at Issue 2 (10 March 2025), transposing Network Code RfG into the GB regime; the limits a connection has to meet are still the same numbers a planner used in 2024.^{4 5} ENA P29 (voltage unbalance, 2 percent at 132 kilovolts and below, up to 1.33 percent allocatable to a single customer) is at Issue 1 from 1990 with a Distribution Code Review Panel revision in progress.⁷ The LTDS published its Stage 2 on 29 May 2026 under the third Ofgem derogation letter of 13 May 2026, which means the network model anyone can plan against is the most current it has ever been.¹

The voltage hierarchy itself has not changed in fifty years. What has changed is the load and generation pattern flowing through it. Solar PV exports back into the LV service push the substation voltage upward during sunny afternoons; heat pumps pull harder on the LV service during winter evenings; EV charging shifts daily peak by an hour or two. The G98 and G99 limits force the network operator to accommodate these flows or to constrain them. The operational answer in 2026 is a combination of network reinforcement (slow), Active Network Management (fast and well understood), and connection-stage flexibility contracts (under active reform).

The statutory frame: Electricity Act 1989 and the GB licence regime

Section 6(1) of the Electricity Act 1989 created the four licence types that still govern the system in May 2026: generation, transmission, distribution, and supply.⁹ A transmission licence authorises operating the assets above 132 kilovolts in England and Wales (and at 132 kilovolts and above in Scotland). A distribution licence authorises operating the assets at and below 132 kilovolts in England and Wales (so the 132 kilovolt tier sits with the DNOs in England and Wales, but with SP Transmission and SSEN Transmission in Scotland; this is the only voltage where the GB regime is asymmetric across the border).

Ofgem grants and amends the licences. Each licence carries Standard Licence Conditions. The condition that produces the Long Term Development Statement on every DNO is SLC 25, and specifically SLC 25.2 which requires publication at intervals of not more than seven years.⁸ A planner reading these notes should follow the chain back: the voltage limit they have to keep inside is in the Electricity Safety, Quality and Continuity Regulations 2002 for the LV service, or in the Engineering Recommendations G98 and G99 for connection-stage and operational-stage limits at higher voltages, and the Engineering Recommendations are part of the Distribution Code which is part of the Distribution Licence. Each layer is a child of the 1989 Act.

The statutory voltage limits at each tier

The headline limit at the LV service is in the Electricity Safety, Quality and Continuity Regulations 2002, regulation 27: the supply voltage shall be 230 volts plus 10 percent or minus 6 percent on a single phase, and 400 volts on three phases with the same envelope. The envelope is asymmetric (more headroom upward than downward) because the network was designed around a 240 volt nominal that aligned upward when the GB regime adopted the 230 volt nominal under harmonisation with the EU in the 1990s; the upward asymmetry preserves the older 240 volt-rated equipment that is still in the field.

For micro-generation and small commercial generators (up to and including 16 amperes per phase), ENA Engineering Recommendation G98 Issue 2 (10 March 2025) sets the connection requirements.⁴ For generators above 16 amperes per phase, up to and including transmission-connected generators (Types A, B, C, D), ENA Engineering Recommendation G99 Issue 2 (10 March 2025) applies.⁵ Both transpose the European Network Code on Requirements for Generators (RfG) into the GB regime.

Two other limits matter at the LV service and are easy to forget. ENA G5/5 (June 2020) sets harmonic compatibility levels up to the 100th order at all voltages; a connection that distorts the supply waveform beyond the G5/5 envelope is non-compliant.⁶ ENA P29 Issue 1 (1990, revision in progress) sets voltage unbalance limits at 2 percent at 132 kV and below, with up to 1.33 percent allocatable to a single customer.⁷ An unbalanced single-phase connection (a large heat pump on one phase only at the end of a long run) can violate P29 before it violates the magnitude envelope.

The per-unit system, base quantities, and why every transmission planner thinks in per-unit

An electrical engineer working on a system with five voltage tiers cannot keep volts and amps in their head. Per-unit normalises everything against a chosen base so the numbers stay close to 1.0 across tiers, and so impedances on different sides of a transformer add directly without a turns-ratio conversion. Per-unit is the working language of network planning.

The base quantities for a three-phase power system are three numbers chosen by the engineer: the base apparent power S_base (in megavolt-amperes), the base line-to-line voltage V_base (in kilovolts), and the base frequency 50 hertz. From these three, the base current, base impedance, and base admittance follow by definition.

The base current I_base is the current that flows when the base apparent power passes through the base voltage on a balanced three-phase system:

I_base = S_base / (√3 · V_base)

The base impedance Z_base is the per-phase impedance that gives a voltage drop of V_base across itself when the base current flows through it:

Z_base = V_base² / S_base

The base admittance Y_base is the reciprocal of Z_base.

Transformer ratios, tap-changing, and the voltage step between tiers

A transformer between tiers has a nameplate ratio (132/33, 33/11, 11/0.4, expressed kilovolts on each side) and an on-load tap-changer that lets the operator move the secondary-side voltage within a tap range, typically minus 10 percent to plus 10 percent in 1.25 percent steps (16 taps either side of nominal plus the central tap, so 33 positions in total).

A 132/33 kilovolt transformer with a nameplate ratio of 132 to 33 has a turns ratio of 4 to 1 ideally. With taps at minus 5 percent on the secondary, the effective secondary voltage at no load is 33 · 0.95 = 31.35 kilovolts; with taps at plus 7.5 percent, it is 33 · 1.075 = 35.475 kilovolts. The tap-changer is the operator's primary tool for moving the downstream voltage profile within the statutory limits as the load and generation pattern changes through the day.

Under heavy load, the voltage drop across the transformer's series impedance plus the cables and lines feeding it pulls the downstream voltage down; the operator raises taps to compensate. Under heavy distributed generation export, the export current rises the voltage; the operator lowers taps. The constraint is the tap range: at the extremes the operator cannot move further, and the choice becomes either reinforcement, curtailment, or a flexibility contract that pays a generator or load to shift.

Voltage drop on a cable or overhead line, with worked examples

Voltage drop ΔV on a line carrying current I through resistance R and reactance X, at a power factor cos φ, is approximately:

ΔV ≈ I · (R · cos φ + X · sin φ) (per phase, in volts)

For most practical purposes the planner treats the cable as a series impedance Z = R + jX in ohms per kilometre, multiplies by the length, multiplies by the current, and reads off the drop. The approximation drops the second-order Ferranti and quadratic terms; for cables less than 50 kilometres long at distribution voltages, the error is below 1 percent and is dominated by the input precision of the impedance value itself.

Standard impedance values (BS EN 50341 for overhead, manufacturer datasheet for cables): a 132 kilovolt overhead line at 240 mm² ACSR is about 0.16 ohms per kilometre resistance plus 0.40 ohms per kilometre reactance; a 33 kilovolt cable at 300 mm² XLPE is about 0.10 ohms per kilometre plus 0.12 ohms per kilometre; an 11 kilovolt cable at 240 mm² XLPE is about 0.13 ohms per kilometre plus 0.10 ohms per kilometre.

Voltage rise from distributed generation and the new constraint pattern

Distributed generation injects power upstream into the network. The flow of current through the network impedance creates a voltage rise at the point of connection compared with the no-generation case. The G98 limit at the point of common coupling for micro-generation is 1 percent steady-state rise; G99 sets parallel limits for larger generators. The rise compounds across multiple generators on the same feeder.

The planner working a new connection at 33 or 11 kilovolts has to compute the rise under the worst-case generation pattern (typically every generator at full output simultaneously, on a low-demand day). When the rise approaches the 6 percent operational envelope minus the existing voltage profile margin, the connection cannot proceed without one of three remedies: reduce the connection size, accept a flexibility contract that curtails the generator during high-rise hours, or reinforce the network. The Connections Reform Gate 2 outcomes in April 2026 (283 gigawatts generation and storage progressed across Phase 1 to 2030 and Phase 2 to 2035) have brought the rise constraint to the front of every connection conversation in 2026, because the queue is now being processed against the actual network rather than against the order it was submitted.¹⁰

Operational voltage control: who does what during the half-hour

Voltage is managed at three timescales and four levels at once. At the transmission level (400 kV and 275 kV), NESO directs the voltage profile through the Grid Code OC1 procedures, using reactive support from generators (synchronous and inverter-based), shunt capacitor and reactor switching at substations, and SVC (Static Var Compensator) and STATCOM devices at strategic nodes.² The half-hourly settlement period is the typical accounting interval; second-by-second voltage management runs much faster than that, with NESO operators viewing system state in real time and acting through commands to the asset owners.

At the 132 kilovolt boundary tier, the transmission owner manages voltage in coordination with NESO. At distribution voltages, the DNO manages voltage through tap-changer position on the bulk supply point and primary substation transformers, plus capacitor banks at strategic substations and (increasingly) Active Network Management schemes that signal connected generators and loads to adjust output. Distribution Code DPC4 covers operational voltage management at distribution.³

Active Network Management is the operational answer to the rise constraint described above. Where a feeder has more headroom under low-export hours than under high-export hours, ANM monitors the local voltage and curtails or releases generator output to keep the voltage inside the envelope, sharing the headroom across multiple connections. The connection contract becomes flexible rather than firm; the generator accepts curtailment in return for a faster, cheaper connection. ANM has been a working pattern on the Orkney Islands since 2009 and has spread across most GB DNOs by 2026.