ENA EREC P28 — voltage fluctuation, explained

What P28 covers

P28 Issue 2 was published in 2018 and became applicable on 23 May 2019, replacing the 1989 Issue 1. It deals with two distinct effects, and it is worth being clear which is which because the assessment routes differ.

Flicker is repetitive voltage fluctuation of sufficient frequency and magnitude to be perceived as instability in lighting. It is characterised by the severity indices P_st and P_lt, measured with a flickermeter conforming to BS EN 61000-4-15.

Rapid voltage changes (RVCs) are fast, non-repetitive transitions between two steady-state voltage levels, typically caused by switching. They are characterised by their peak excursion and their settled change, and grouped into categories by how often they occur.

P28 does not cover harmonic voltage distortion — that is G5 — nor voltage unbalance, which is P29. It applies to new connections and to existing connections being modified; connections previously assessed under Issue 1 are not retrospectively reassessed.

The 3% step voltage change limit

Clause 5.4 sets a general limit of 3% on the magnitude of a percentage step voltage change, expressed against the nominal system voltage and measured at the point of common coupling. It applies regardless of how often the change occurs, and Clause 6.2 makes conformance to it the first thing to assess.

Two practical points. The limit applies at the PCC, not at the equipment terminals — measured at the terminals the change can legitimately be larger, and BS EN 61000-3-3 and 61000-3-11 state 3.3% there. And Clause 6.2 gives the network operator discretion to permit larger changes in special circumstances, the example given being a large motor started only once in several months.

Flicker: P_st, P_lt and planning levels

P_st is short-term flicker severity, derived from a ten-minute measurement window. P_lt is long-term severity, derived from twelve consecutive P_st values — a two-hour window — by a cubic mean.

Planning levels are the values a network operator assesses new connections against. They sit at or below the compatibility levels, and are lower at higher voltages because flicker transfers downward through the network.

Table 2. LV planning levels equal the compatibility levels in Table 1.

The number most engineers actually need is different, though. For assessing a single item of equipment, the operative threshold is the Stage 2 limit of P_st ≤ 0.5 at the PCC (Clause 6.3.3). Equipment at or below it connects without further detailed assessment. The margin between 0.5 and the planning level of 1.0 exists because disturbances summate: eight separate loads each at P_st = 0.5, combined with a cubic summation law, reach exactly 1.0.

Rapid voltage changes and the three categories

RVC limits are not a single number but a time-dependent envelope, and which envelope applies depends on how often the event occurs. The categories are set out in Table 4 and drawn in Figures 5 to 7.

Table 4 and its notes. Steady-state change is limited to 3% in every category.

Categories 2 and 3 carry conditions beyond the envelope, and they are easy to miss:

• No more than 1 event per day.
• An event consists of no more than 4 rapid voltage changes.
• Each rapid voltage change separated by at least 10 minutes.
• All switching completed within a 2-hour window.

Category 2 events need no notification. Planned Category 3 events must be notified in advance to the Transmission System Operator at transmission voltages; at distribution voltages notification is at the network operator's discretion. In Northern Ireland, tighter limits than Figure 7 apply while G59/1/NI is in force.

The three-stage assessment

Flicker assessment follows three stages, summarised in Figure 9. The important thing to understand is that this is a flowchart, not three tests you run in parallel — satisfying an earlier stage ends the assessment.

• Stage 1 — a simplified assessment for discrete items of LV equipment, based on conformance to a BS EN 61000 product standard. Equipment meeting it connects without reference to the network operator or further assessment. It does not apply to HV connections, nor where the PCC is at HV, nor to equipment that works together as a system.
• Stage 2 — calculate the worst-case P_st at the PCC under the worst-case normal operating condition. At or below 0.5, connect without further assessment. No measurement of background flicker is required.
• Stage 3 — for equipment failing Stage 2, and only with the network operator's agreement. Existing background flicker is measured, combined severity is calculated, and the available headroom decides the outcome.

Stage 1 thresholds

The Stage 1 routes for LV equipment, from Clauses 6.3.2.1 and 6.3.2.2. Rated current is per phase.

The P_st = 0.5 characteristic

For Stage 2, Clause 6.3.3.2 gives a simplified route: plot your step voltage change against the minimum interval between changes, and if the point sits on or below the P_st = 0.5 curve of Figure B.1.2, the equipment connects without further detailed assessment.

The curve is derived from the P_st = 1 characteristic in PD IEC/TR 61000-3-7 by halving the voltage axis, since P_st is linear in voltage-change magnitude. It is deliberately capped at 3% once every 475 seconds.

One more rule worth knowing: where a calculated result lands within ±10% of the limit, the standard directs that flicker severity be assessed by simulation rather than relied on from the simplified method. Our calculator returns those as borderline rather than as a bare pass.

Generic supply impedances

Where no measurement or network-operator data is available, Table 7 gives generic LV supply impedances for approximate calculation. For a declared supply capacity of 100 A or more, the standard directs that operator-specific data be used instead.

Table 7. Actual impedances can be higher — a pole-mounted transformer with a low rating, or a small cross-section service cable, will both exceed these.

Motors: Annex A

Annex A lists small LV motors that can be connected direct-on-line without consideration of flicker or RVC. Which table applies depends on how often the motor starts and where the point of common coupling is.

Where the PCC is the LV busbar of a distribution substation and the motor starts at intervals of ten minutes or longer, Table A.1.3 permits considerably more, scaled to the transformer:

Clause A.2 allows 1.5× these ratings where star-delta starting is used. Note that a reduced-voltage starter causes a second voltage change at changeover, which counts as a further step change in its own right.

The equations

For a balanced three-phase supply, the percentage voltage change caused by disturbing equipment comes from Clause 6.3.5.

Where supply impedance is given as per-unit resistance and reactance on a base MVA (Equation 3):

ΔV/V = (S / S_base) × (cos φ · R_p.u. + sin φ · X_p.u.)

Where the supply short-circuit power is known in MVA, and the load power factor is assumed equal to the ratio of supply resistance to supply impedance — the worst case — it simplifies to Equation 4, which is what most assessments use:

%ΔV = (S / S_sc) × 100

Flicker from multiple sources combines by a summation law (Equation 1), with the exponent chosen for how likely the fluctuations are to coincide: 4 where coincidence is very unlikely, 3 as the general default, 2 where coincidence is likely, and 1 where it is near-certain, such as several motors started together.

Equipment-specific guidance

Clause 8 gives specific treatment for particular equipment types. The ones that come up most:

• EV charging equipment (8.6). Fixed charging equipment connects under Stage 1 at ≤ 16 A with BS EN 61000-3-3 conformance, or at ≤ 32 A if installed at a domestic residence. Above that, conditional connection under 61000-3-11 applies. Network operators are directed to give special consideration to multiple chargers at one PCC, including preventing simultaneous switching.
• Heat pumps (8.5). A single domestic heat pump follows the ENA notification process, whose requirements are equivalent to Stage 1. Indoor and outdoor parts must be tested as one integrated system as well as individually. Direct-on-line heat pumps warrant special consideration.
• Motors (8.3). For direct-on-line starting the whole cycle may be treated as one step change. Where the front time is short and the tail comparatively long, the maximum voltage change can be used directly. Manufacturer locked-rotor data is preferred over assumption.
• Wind, PV and storage (8.7–8.9), welding (8.11) and furnaces (8.4) each have their own treatment and generally need a full study rather than a simplified assessment.