The DNO Fuse Limitation: Managing the 60A to 100A Service Head Bottleneck for Multi-Asset Homes

The Supply Boundary as the Binding Constraint

The conventional narrative around residential decarbonisation concentrates on the assets themselves: the efficiency of the heat pump, the capacity of the battery, the charging rate of the EV. A less visible constraint sits upstream of all of them, at the point where the public distribution network meets the private installation. This boundary, defined physically by the Distribution Network Operator (DNO) cutout and the service cable that feeds it, was engineered in an era when a household's maximum demand was dominated by a kettle, an electric cooker and perhaps an immersion heater, and when the statistical assumption was that these loads would rarely coincide. The 60 A, 80 A and 100 A single-phase service heads that still serve the majority of British homes embody that assumption. They were never designed for a household in which an EV charger, a heat pump and a battery inverter can each legitimately demand several kilowatts for hours at a time.

This audit characterises the engineering of the cutout assembly, quantifies the over-allocation arithmetic of a multi-asset home, maps the civil and regulatory friction involved in upgrading a looped service, and deconstructs the control mechanics of the current transformer (CT) based load management that has become the standard workaround. It is written for electrical contractors, retrofit coordinators and asset planners who must decide, property by property, whether to pursue a network upgrade or to engineer within the existing envelope.

Cutout Assembly Engineering and Fuse Physics

The cutout, formally the DNO's service termination and main protective device, is the first element of the consumer's electrical installation and the last element of the network. In its traditional form it is a cast housing, historically filled with bituminous compound (pitch) to provide insulation and moisture exclusion around the cable termination, containing the incoming service conductor termination, a neutral link and a single-pole fuse carrier on the line conductor. Modern installations use thermoplastic enclosures that replace the pitch fill with a dry, mechanically retained termination and offer a defined ingress protection rating, improved fire performance and simpler isolation. Both configurations are sealed by the DNO, and the consumer has no right to open or alter them.

The fuse inside the carrier is a cartridge element manufactured to BS 1361 (cartridge fuses for AC circuits in domestic and similar premises) or to BS 88-3 (the part of the BS 88 series covering fuses for use by authorised persons, including the supply-side cutout application). The element consists of a silver or copper strip with calibrated notches, supported within a ceramic body and surrounded by a quartz sand filling. In normal operation the strip dissipates a small amount of heat, I squared R, and this is shed through the body, the end caps and the carrier into the surrounding enclosure. When a fault current flows, the notched sections reach the melting point almost instantly, an arc forms across the gap, and the sand absorbs the arc energy by fusing into a high-resistance silicate, thereby extinguishing the current. The breaking capacity of these devices, commonly 16 kA or more, is a key reason they are preferred at the service head, where prospective fault currents can be substantial.

Time-Current Characteristics and the Thermal Reality

A fuse is not a threshold device. It is an inverse-time device defined by a time-current characteristic curve that plots the pre-arcing time against the prospective current on logarithmic axes. At a multiple of the rated current of several times In, a fuse will operate in milliseconds. At modest overloads, in the region of 1.25 to 1.6 times In, the curve flattens: the conventional non-fusing current, below which the fuse must carry the load for the conventional time (typically one hour or more for this rating range), is approximately 1.25 In, and the conventional fusing current, at which it must operate within that time, is approximately 1.6 In. A 100 A cutout fuse may therefore continue to carry 125 A for an extended period without necessarily operating, and may take a long time to operate at 150 A. The fuse is thus an excellent short-circuit device and a relatively blunt thermal overload device. Its role in protecting the service cable relies on the cable's own thermal capacity being matched to that behaviour.

The critical, and widely misunderstood, issue is not whether a continuous load will cause the fuse to blow in the short term, but how it affects long-term thermal degradation. A fuse carrying a high proportion of its rating continuously, for the six, eight or ten hours of an EV charging session or a heat pump heating cycle, runs materially hotter than one subject to the intermittent loads for which the system was originally diversified. Elevated temperature accelerates ageing of the element: repeated thermal cycling causes grain growth and fatigue at the notches, gradually changing the element's resistance and shifting its time-current curve. Contact resistance in the carrier and at the terminations increases with oxidation and thermal relaxation of the clamping force. Higher resistance causes higher local heating, which accelerates oxidation, creating a positive feedback loop that, in a worst case, produces a hot joint, a discoloured or melted carrier, or a fire risk. This is the physical basis for the DNO's caution about sustained loads at or near the rating of the cutout, and the reason that continuous-duty assets must be assessed differently from intermittent domestic appliances.

The ambient environment compounds the issue. Cutouts are frequently located in enclosed external boxes, in under-stairs cupboards, or in meter cupboards with limited ventilation and often adjacent to the meter, the consumer unit and any added equipment. Derating for ambient temperature, grouping and enclosure type is rarely considered by those who focus on the nominal fuse rating alone, yet in practice it reduces the sustained current that the assembly can carry without exceeding its design temperature rise.

The Multi-Asset Over-Allocation Equation

The historical basis for domestic supply sizing is the After Diversity Maximum Demand (ADMD), a statistical figure representing the average maximum demand per dwelling when a large group of dwellings is considered together. Traditional ADMD values were of the order of 2 kW per property, with higher figures where electric heating was installed. The assumption is that individual loads switch on and off randomly, and that the likelihood of all of them operating simultaneously is vanishingly small. Low-carbon technologies break this assumption in two ways. First, they are high-power and long-duration. Second, they are correlated: EV charging begins when the owner arrives home, heat pumps operate during cold evenings, and both coincide with cooking and lighting peaks. The diversity that justified the original sizing disappears.

Consider a representative multi-asset home and sum its simultaneous maximum single-phase demands at a nominal 230 V. The EV supply equipment (EVSE) is configured for 32 A continuous, equivalent to 7.36 kW. A heat pump with a 16 A start-up spike, being the peak across the compressor and associated auxiliaries before the unit settles to a lower running current. An induction hob drawing 13 A at a typical multi-zone cooking load, approximately 3 kW. A household baseload of lighting, refrigeration, electronics and appliances, taken conservatively at 1.5 kW, or approximately 6.5 A.

The arithmetic is direct. The sum of 32 A, 16 A, 13 A and 6.5 A is 67.5 A. A 60 A cutout, with a continuous power capacity of 13.8 kW (60 A multiplied by 230 V), is therefore overloaded by 7.5 A, or 12.5%, in this plausible and unexceptional scenario. That overload will not necessarily blow the fuse immediately, as the time-current curve indicates, but it places the assembly in a thermal regime outside its design basis, and in the case of a continuous EV charge in combination with an evening heat pump cycle, it may be sustained for hours. An 80 A cutout, with an 18.4 kW capacity, accommodates the 67.5 A total with a margin of 12.5 A, but only just. The introduction of a further 13 A immersion heater or 3 kW of resistive backup heating, the typical heat pump auxiliary element, brings the total to 80.5 A and exceeds the 80 A rating. A hybrid solar battery inverter configured to charge from the grid at 3.6 kW adds approximately 15.7 A, giving a total of approximately 96 A and leaving a 100 A cutout, with a 23.0 kW capacity, with a margin of only 4 A.

These values are non-diversified, and critics will correctly observe that all of these loads rarely peak simultaneously. That observation is precisely why the risk is difficult to dismiss and difficult to quantify: the probability of coincidence is low per day but not negligible over a year, and the consequence is concentrated at the shared boundary asset. Moreover, the habits of the occupants, the tariffs they adopt and the control strategies they use can all increase correlation. A smart tariff that rewards charging between 00:30 and 05:30 will concentrate EV charging, battery charging and any scheduled hot water cycles into the same window, a form of engineered coincidence. A diversity model calibrated to behaviour that existed before smart tariffs is no longer sufficient.

The curtailment thresholds are expressed as indicative engineering practice rather than as a regulatory requirement. Installers typically set the load management limit below the fuse rating to retain headroom for the response time of the control loop, for measurement tolerance in the CT and for uncontrolled loads that are not subject to curtailment. The conservative approach is to size the limit against the lowest-rated element in the chain, which may be the service cable or the meter tails rather than the fuse.

The Civic Friction of Unlooping and Feeder Cable Upgrades

When a property owner applies to the DNO to increase the capacity of the supply, the first technical question is not the fuse but the cable that feeds it. In much of the UK's older housing stock, particularly in terraced streets and semi-detached properties, services were installed in a looped arrangement: a single underground cable runs from the main in the street to the first property, where it is jointed, and a continuation runs on to the neighbouring property. The two or more properties share a single service cable and, in some cases, a single section of the street main connection. The rating of that shared cable is determined by its construction, its age and its installation conditions, and it was sized for the combined diversified demand of the properties it serves, not for the non-diversified demand of a fully electrified home.

When a DNO assessor identifies a looped service in response to an application for a capacity increase, the scope of the work changes categorically. Replacing or upgrading the cutout fuse alone would increase the available current at the property without increasing the capacity of the shared cable, which would then be at risk of overload, particularly if neighbouring properties are also adding load. The DNO must therefore either unloop the service, giving each property its own dedicated cable from the main, or replace the service with a higher-capacity cable. Both options require street-level excavation.

Execution Processes Across the Major DNOs

The principal operators, including UK Power Networks, National Grid Electricity Distribution, Scottish Power Energy Networks, Northern Powergrid, Electricity North West, and SSEN Distribution, each publish their own connection application processes, charging statements and indicative lead times, but the underlying sequence is similar. The applicant submits a request for a capacity increase through the DNO's online portal, supplying information on the intended loads, including the EV charger, heat pump and any generation or storage. The DNO assesses the existing supply against the new maximum demand, considering the service cable, the local mains, the substation transformer and the cumulative effect on the low-voltage network. Where reinforcement is needed, the DNO issues a quotation that distinguishes between work that is contestable, which may be carried out by an accredited independent connection provider, and work that is non-contestable and must be performed by the DNO itself.

The execution of a service upgrade involves several dependent activities. A site survey and cable trace identify the route and construction of the existing service. A design is prepared and, if the work occurs on the public highway or footway, an application is made to the local highway authority. Under the New Roads and Street Works Act 1991 and the permit schemes introduced under the Traffic Management Act 2004, utility works in the public highway require prior notification or a permit, with conditions governing the duration, traffic management, reinstatement and noise. In permit-scheme areas the application is subject to the authority's review and may be refused or conditioned, and the programme may be influenced by other planned works, seasonal restrictions and permit coordination. Where the work affects a footway across a driveway or a shared private access, the consent of other owners may also be required.

On site, the DNO's jointers and excavation teams isolate the supply, expose the existing cable, install the new service from the main to the property, and make the jointing and termination. Jointing is a specialised skill and each joint is a potential point of failure, so the process includes testing, documentation and often a requirement for specific jointer approvals. The unlooping of one property requires a temporary or permanent supply interruption to the neighbouring property that shares the cable, creating a need for notification, coordination and sometimes temporary generation or phased outages. The new service is then terminated at a new cutout, the metering is reconnected by the meter operator and the permanent reinstatement of the footway or carriageway is completed, sometimes after an interim temporary reinstatement and a subsequent permanent restoration visit.

The consequence for the retrofit programme is a combination of cost uncertainty and programme risk. Charges for upgrades can range from negligible, where only a cutout fuse change is needed, to several thousand pounds where significant excavation is required, and lead times can range from a few weeks to many months depending on the DNO's workload, the permit process and the complexity of the street. For a retrofit coordinator managing a heat pump installation and a wider project timeline, the DNO is frequently the critical path item. It is an external dependency that cannot be compressed by additional contractor resource, and it is the principal reason that load management, rather than reinforcement, has become the default strategy for an increasing proportion of installations.

Dynamic Load Balancing and CT Clamping Mechanics

Dynamic load management transforms the supply constraint from a physical limit that must be removed into a control variable that is managed in real time. The principle is simple: measure the total current drawn by the property at the supply boundary, compare it with a configured limit, and reduce the power of the flexible loads, usually the EV charger, when the limit is approached. The implementation, however, depends on the quality of the measurement and the speed and robustness of the control loop.

Measurement: The Split-Core Current Transformer

The measurement element is a split-core current transformer clipped around the primary phase conductor, typically the 25 mm squared meter tail (or the equivalent main tail between the meter and consumer unit). The principle is that of a transformer in which the conductor carrying the load current is the single-turn primary, and a secondary winding of many turns on the clamp's core produces a proportionally smaller current or voltage. Alternating current in the conductor creates an alternating magnetic flux in the core. That flux induces a current in the secondary winding that is proportional to the primary current, at a ratio defined by the turns ratio, so that, for example, 100 A in the conductor may produce 50 mA or a defined millivolt signal at the secondary. The device therefore measures the cumulative effect of everything downstream of the clamp, as the magnetic flux is the vector sum of all currents flowing through the conductor.

The location of the clamp is critical. It must be installed on the line conductor upstream of all controlled and uncontrolled loads for which the supply boundary is to be protected, which means on the meter tail on the load side of the meter or at the incomer of the consumer unit. The clamp must be oriented in the correct direction, as reversed orientation produces a negative reading and, in installations with generation, confuses import and export. It must enclose only the single line conductor, since clamping both line and neutral produces a net zero reading as the fluxes cancel. Installation on the meter tails requires the work to be performed by a competent person and may require the cooperation of the energy supplier or meter operator, where seals protect the metering side. Accuracy depends on the core quality, the burden resistor value, the closure of the split core and the absence of an air gap, and typical clamps deliver accuracy in the range of 1% to 2% when correctly installed, with errors increasing at low currents.

Control: From Flux to Duty Cycle

The signal from the CT is conditioned and digitised, either in the EVSE itself or in a separate energy monitor or gateway. The processor calculates the root-mean-square current, subtracts the contribution of the controlled load to estimate the uncontrolled household demand, and determines the headroom available to the controlled load under the configured limit. For a 60 A cutout with a 50 A limit and a household demand of 28 A, the headroom for the EVSE is 22 A. The controller sets the EVSE current limit accordingly. In a Mode 3 charging system, this is achieved by changing the duty cycle of the 1 kHz pulse-width modulated control pilot signal, which tells the vehicle the maximum current it may draw: a limit of 22 A corresponds to a duty cycle of approximately 36.7%, derived from the standard relationship of 0.6 A per percentage point of duty cycle. The vehicle's on-board charger responds by reducing its current draw, typically within a few seconds, though the standard permits the vehicle some time to comply.

Communication between the measurement device and the EVSE is accomplished through a local bus or network. Common implementations use Modbus over RS-485 or TCP, proprietary wired or wireless protocols, or integration through a manufacturer's own energy management ecosystem. The sub-second update interval of the measurement and communication handshake is important: if the household load rises abruptly, for example when an induction hob and a kettle are switched on in quick succession, the total current may exceed the limit within the control interval, and the system must reduce the EVSE current before the thermal inertia of the fuse and cable is exceeded. Because fuses and cables respond thermally over minutes rather than milliseconds, a brief excursion of a few seconds is not damaging, which is why control intervals of one second or so are considered adequate, provided the loop is stable and the response is monotonic.

Failure Modes and Fail-Safe Design

The robustness of the entire strategy depends on the behaviour of the system when communication or measurement fails. A load management system that loses its CT signal, whether through a disconnected clamp, a failed bus or a loss of wireless link, must default to a safe condition rather than to full power. Well-designed equipment reduces the EVSE to a pre-configured fail-safe current, typically 6 A, which is the protocol minimum, or suspends charging altogether. Installers should verify this behaviour during commissioning by deliberately disconnecting the CT and confirming the response. Other risks include incorrect CT ratio configuration, which results in the system believing the load to be lower or higher than reality, and clamp placement downstream of loads that are not intended to be excluded. Sizing the limit conservatively and documenting the configuration, the clamp location and the fail-safe behaviour are an essential part of the handover pack.

It is also necessary to recognise the limits of what load management can protect. It governs the controlled loads only. An uncontrolled load, such as a cooker, a shower or an immersion heater, that independently approaches the supply limit cannot be curtailed by the EVSE alone. A properly designed system therefore reserves headroom for the maximum uncontrolled demand or extends the control to additional loads, such as the heat pump's resistive backup, the immersion heater or the battery inverter's grid-charging rate. In installations with generation and storage, the DNO may additionally require export limiting in accordance with the relevant ENA engineering recommendations, and the load management and export limitation systems must be designed to work together rather than in conflict.

The Cost of Curtailment and Asset-Optimisation Implications

Dynamic curtailment is not a free alternative to reinforcement. Its cost is performance: during the evening peak, when the household's other loads are highest, the EV charger is throttled, and in the extreme to the 6 A minimum or to a full suspension. For most households this cost is modest because the EV is typically plugged in for far longer than the time it needs to charge. A vehicle requiring 30 kWh can be replenished at 3.7 kW in about eight hours, comfortably within an overnight window. The relevant question for an asset planner is therefore not whether curtailment occurs but whether the cumulative energy delivered within the available time window meets the occupant's needs, which can be modelled using the household's demand profile and the charge-time requirement.

Where the vehicle usage is high, the tariff window is narrow, or several assets compete in the same period, curtailment may become binding. In these cases, the order of priority between assets becomes a design decision: heating and hot water are normally treated as essential loads, while EV charging and battery charging are flexible. Specifiers should document the priority hierarchy, communicate the performance implications to the client, and ensure that the tariff and scheduling strategy are consistent with the physical limits of the supply.

The decision framework that follows from this analysis is as follows. Where the cutout is rated at 100 A on a dedicated service, load management is generally sufficient and a network upgrade is rarely justified. Where the cutout is rated at 80 A, the decision turns on the asset mix: a modest combination of EV charging and heat pump can usually be accommodated with a CT-based limit, whereas a larger set of assets suggests an application to the DNO. Where the cutout is rated at 60 A, an application for upgrade should be made early, because lead times are long, but load management should still be commissioned as an interim or permanent measure to protect the existing infrastructure. In every case the correct sequence is to notify or apply to the DNO in line with the relevant guidance, to design the control strategy around the weakest element in the supply chain, to commission and test the fail-safe behaviour, and to provide the client with a clear account of the trade-off between cost, programme and performance. Treated in this way, the cutout ceases to be an unmanaged risk and becomes a defined design parameter in the optimisation of a multi-asset home.