Unlocking Grid Capacity Without Building New Infrastructure

The electricity grid is running out of room. Across the industrialised world, the wait to secure a new grid connection, or to upgrade an existing one, has grown from months to years.

The causes are well documented: surging demand from data centres, the electrification of transport and heating, the connection of renewable generation, and decades of underinvestment in the grid. The conventional response, more substations, more cables and bigger transformers, is necessary but painfully slow. New transmission lines take many years from planning to energisation. Distribution reinforcement is faster, but still measured in years, not months.

What is far less discussed is the capacity that is already connected but not doing work. At many industrial sites, reactive power and harmonic distortion take up part of the grid connection without delivering anything useful. That capacity is paid for, connected and physically present. It is simply being used inefficiently.

This article examines how correcting power quality frees apparent power within an existing grid connection, letting a site lower its agreed capacity or make room for the electrification that decarbonisation demands without an upgrade, and where that freed capacity stops.

01The capacity crunch

The scale of the connection backlog is hard to overstate. It is one of the most significant bottlenecks to industrial growth, electrification and decarbonisation in the developed world.

In Great Britain, the queue of projects waiting to connect to the transmission network passed 700 GW in 2024, most of it generation and storage, and National Grid ESO has offered some projects connection dates in the late 2030s. Distribution networks are under similar pressure: new commercial and industrial connections often need reinforcement works that add years and substantial cost.

In Germany, the Energiewende depends on new transmission lines to carry northern wind power to the industrial south, and the largest of them are running years behind their original dates. Distribution networks face a rising tide of connection requests from heat pumps, EV chargers and industrial electrification at once.

In the United States, Lawrence Berkeley National Laboratory counted nearly 2,600 GW of proposed capacity waiting in the queues for transmission interconnection at the end of 2023. The typical project built that year took nearly five years from its interconnection request to commercial operation, against three years in 2015. The queues are dominated by solar, wind and battery storage, but load growth, much of it from data centres, is adding to the pressure.

Globally, the International Energy Agency counts at least 3,000 GW of renewable projects waiting in grid connection queues. It estimates that 80 million kilometres of grid must be added or replaced by 2040, as much as exists in the world today, and that grid investment needs to nearly double, to over USD 600 billion a year by 2030.

Of renewable projects waiting in grid connection queues worldwide (IEA, 2023)

3,000+ GW

Kilometres of grid to add or replace by 2040 (IEA)

80 million

A year of grid investment needed by 2030, nearly double today’s (IEA)

$600bn+

The bottleneck is increasingly not generation but the network that connects generation to demand. For an industrial facility that needs more capacity, to install EV charging, electrify a gas-fired process or add a production line, the message from network operators is increasingly the same: wait. Reinforcement is expensive, timelines are long, and the queue grows faster than it clears.

But what if some of the capacity is already there?

02The hidden capacity in your existing connection

A grid connection is physically rated in kilovolt-amperes, kVA: apparent power, not real power. A site fed through a 2,000 kVA transformer has 2,000 kVA of capacity to draw on. How much useful power that connection delivers depends on the site’s power factor.

Power factor is the ratio of real power, in kW, to apparent power, in kVA. At 1.0, every amp in the connection does useful work. Below it, reactive power, which sustains the magnetic fields in motors and transformers and contributes nothing to output, takes up part of the connection. Real and reactive power add at right angles, not end to end: at a power factor of 0.80, a site’s 1,600 kW of real power travels with 1,200 kvar of reactive power, and together they fill 2,000 kVA.

The arithmetic is straightforward. A 2,000 kVA connection at a power factor of 1.0 can deliver 2,000 kW of useful power; corrected to 0.98, it delivers 1,960 kW. At 0.85 it delivers only 1,700 kW, at 0.80 just 1,600 kW, and at 0.70 just 1,400 kW.

That 400 kW gap at a power factor of 0.80 is not theoretical. It is real capacity, in cables, transformer and switchgear, occupied by current that does no work. The cables are heating. The transformer is loaded. The protection is measuring current. Yet the current is a quarter larger than the work alone needs: 1 ÷ 0.80 = 1.25.

Industrial power factor is often poor. Induction motors, which dominate industrial load, run at their best power factor near full load, and it falls steeply as the load drops, as IEEE Std 141, the Red Book, sets out. Most motors are sized for peak duty and spend much of their time well below it, so site power factors below 0.85 are common where nothing has been corrected.

The implication is clear. At the power factors common on uncorrected sites, a good part of a connection carries current that does no work. Correcting to 0.98+ frees that apparent power for the site’s own use: the equivalent of a connection upgrade, achieved without new cables or transformers and without a place in any queue, wherever the connection’s limit is set in kVA. Section 04 works an example through, and says what changes where the limit is set in kW.

03How harmonics compound the problem

Reactive power is not the only thing taking up capacity. Harmonic distortion, currents at multiples of the 50 or 60 Hz fundamental, adds a further burden, and an often underappreciated one.

Harmonic currents come from non-linear loads: variable-frequency drives, rectifiers, LED lighting drivers, uninterruptible power supplies, arc furnaces and switched-mode power supplies. They draw current in sharp pulses rather than smooth sine waves, injecting harmonics, typically the 3rd, 5th, 7th, 11th and 13th, into the site’s network.

The problem is that harmonic currents heat electrical infrastructure out of proportion to their size. In a transformer they raise the winding eddy-current and stray losses, and IEEE C57.110-2018 sets out the standard method for accounting for them. It weights the eddy-current loss from each harmonic by the square of its order, so, amp for amp, the 5th harmonic heats the windings 25 times as much as the fundamental, and the 7th 49 times.

The practical consequence is transformer derating. A transformer carrying significant harmonic current cannot safely deliver its full nameplate rating, because the extra heat uses up its thermal headroom. How far it must be derated depends on the harmonic spectrum and on the transformer’s own design. In the worked example in our article on harmonics, a drive-dominated load at about 25% current THD leaves a standard 1,000 kVA transformer good for about 900 kVA.

Harmonics add losses in cables too. Skin and proximity effects raise a conductor’s resistance at higher frequencies, so harmonic current makes more heat per amp than the fundamental. And triplen harmonics, the 3rd and its odd multiples, add up in the neutral instead of cancelling: IEC 60364-5-52 applies a reduction factor of 0.86 to a four- or five-core cable once the third harmonic passes 15% of the phase current, and above 33% sizes the cable on the current in its neutral.

The result is that harmonics and reactive power compound the capacity problem. A facility with both a poor power factor and high harmonic distortion is losing capacity to two separate mechanisms at once.

04The maths of freeing capacity

To make the scale of the opportunity concrete, take a worked example on a hypothetical site.

Worked exampleA manufacturing site has a 3,000 kVA grid connection, agreed in kVA, and at its peak it uses all of it. Its meter shows a displacement power factor of 0.82 and 18% current THD. It wants to add about 500 kW of new load, an electrified heating process and a bank of vehicle chargers, and its distribution network operator has told it that this needs a connection upgrade, with a lead time measured in years.

Here is the calculation, step by step.

1. The true power factor Harmonic current adds apparent power that the displacement figure does not show. The true power factor counts both, and with low voltage distortion it is close to the displacement power factor ÷ √(1 + THD²): here 0.82 ÷ √(1 + 0.18²) = 0.807.
2. The real power the connection carries today At its 3,000 kVA limit, the site draws 3,000 × 0.807 = 2,421 kW of real power. The rest of the connection carries reactive and harmonic current, through the transformer and every cable, doing no work.
3. The same load, corrected Corrected to a true power factor of 0.98, the same 2,421 kW needs 2,421 ÷ 0.98 = 2,470 kVA. That frees 530 kVA of the connection.
4. What the connection can carry now At 0.98, 3,000 kVA carries 3,000 × 0.98 = 2,940 kW: room for 519 kW of new load at the same power factor, enough for the 500 kW the site wants to add.

The upgrade the site was told it needed is no longer required for this load. It meets its electrification plans within the connection it already has, provided the new load runs at the same power factor.

The transformer gains thermal margin too. If the site’s own transformer has been derated for its harmonic load under IEEE C57.110, removing the harmonic current gives that margin back. How much depends on the spectrum and on the transformer, so it is measured on site, not assumed.

Which capacity the correction frees

What the 530 kVA is worth depends on what limits the site, and on how its capacity is agreed with the network operator: in kVA or in kW.

Inside the site boundary, the freed apparent power is real headroom whatever the tariff. Transformers, cables and switchgear are rated in kVA or amps, so a site that corrects its power quality can defer an upgrade of its own equipment even where its connection is agreed in kW.

05Real-world implications

The worked example is not unusual. Across many industrial sectors the pattern repeats: sites that believe they have used up their grid connection are in fact spending part of it on reactive power and harmonics. Freed capacity has practical uses across the range of industrial electrification.

EV charging infrastructure

Workplace and depot charging is one of the fastest-growing sources of new load. A fleet of electric vans can need several hundred kilowatts of charging capacity, enough to push many connections to their limit or past it. Where capacity is agreed in kVA, freed apparent power can make room for some or all of it without an upgrade.

Process electrification

Replacing gas-fired heating, drying and thermal processes with electric ones is central to industrial decarbonisation. Heat pumps, infrared and induction heating and electric boilers all need significant electrical capacity, and some of it can often be found within the existing connection.

Production expansion

New production lines, more compressors, more automation: any expansion needs electrical capacity. The alternative to freeing it is a connection upgrade: an application to the distribution network operator, a wait for the design study, a contribution to reinforcement works, and a construction period that may constrain production. Capacity freed within the existing connection avoids all of it.

On-site generation and storage

Battery storage and on-site solar both depend on connection capacity. A battery that charges at 500 kW needs 500 kW of import headroom while it charges, which freed apparent power can provide where import capacity is agreed in kVA. Export capacity is agreed separately, and correcting the site’s load does not change it.

The time advantage

A connection upgrade takes years from application to energisation, through design, consents, procurement and construction. Correcting power quality inside the site boundary does not wait for network reinforcement. For facilities with urgent electrification timelines, driven by fleet decarbonisation targets, process deadlines or carbon reporting, that difference can be decisive.

06The grid-level benefit

The benefits of correcting power quality are not limited to the site that does it. When a site draws less reactive and harmonic current, that current no longer flows through the network operator’s cables and transformers either. Distribution transformers run cooler, network losses fall, and less distortion reaches the other sites on the same network. Across the industrial base, that adds up.

It is worth being precise about what this capacity is. Freed apparent power is headroom in a site’s own connection and equipment, and relief for the network assets that feed it. It is not new generation, and where a network operator agrees capacity in kW, it does not become more kilowatts a site may draw.

The build-out the IEA calls for is vast. Capacity used better inside existing connections cannot replace that investment, but it can ease the pressure on it, connection by connection. This is not a zero-sum intervention. It is a positive-sum one: the site that corrects its power quality frees capacity for its own use, lowers its own costs and reduces the burden it puts on the shared network.

The demand-side role in grid planning

Network operators increasingly recognise that demand-side measures can relieve network constraints. In Great Britain, Ofgem’s RIIO-ED2 price control encourages distribution network operators to use flexibility and other alternatives to building new capacity where they cost less than reinforcement. Power factor correction is not a flexibility service in that sense, but it works in the same direction: it lowers the current a site draws from the network without lowering its output.

Connection terms can already set a minimum power factor, and a site that draws little reactive current asks less of the network than one doing the same work at a poor power factor. As capacity tightens, efficient use of an existing connection is likely to be valued alongside new construction.

For industrial facilities, that is both an opportunity and an imperative. The opportunity is to free capacity quickly, without waiting for reinforcement. The imperative is that networks are becoming less able to accommodate inefficient loads. Sites that deal with their power quality early will find it easier to secure capacity and negotiate connection terms. Those that do not may find themselves at the back of a long queue.

A bridge to the grid of the future

The grid reinforcement the world needs will take decades. In the meantime, every kilovolt-ampere freed within an existing connection can be put to work: charging vehicles, electrifying processes, supporting on-site renewables and enabling the growth the economy demands.

Correcting power quality is not a substitute for grid investment. The long-term build-out of transmission and distribution must accelerate. But it is a bridge: a way to create headroom now, within infrastructure that already exists, while the longer-term solutions are planned, financed and built.

The capacity is there, inside the connection. The question is whether it is put to productive use, or left to reactive current and harmonics that serve no purpose at all.