Renewables and network integration issues, with the UK as an example

Guest blog by David B Watson BSc C.Eng FIES MIET.

Many nations are working to increase electrical generation from renewable sources such as wind, solar, tide and wave. The majority of the public support the principle of decreasing generation from sources that emit greenhouse gases, namely gas, coal and oil.

In most wealthy countries the energy industry comprises private generating and distributing companies where the mix of energy generation is sought to be influenced/determined by the levers remaining in the control of elected governments. These can typically include subsidies, grants, varying Contracts for Difference and constraint payments applicable to favoured technologies and financial models for fast responding reserve and Capacity Market/back up from less favoured and/or less deployed technology such as CCGT, diesel generation, OCGT and coal.

This may imply a tautological truth that renewables are inexorably, seamlessly and dependably replacing fossil fuel-based electricity supplies within national grids.

This is not the case. There are harsh prevailing and unsolved challenges. There are consequences to the increase in renewables generation and penetration as it adversely affects the support of power quality and as the ability of existing transmission grids to cope diminishes.

Key among the challenges are:

  • Intermittency of renewable generation affects and even risks grid stability. It challenges the gas grid rapid “ramp up” capability to supply increasing back up generation as installed renewables and the resultant power failure “swing” magnitudes increase. In the UK, Centrica have warned that the gas grid will reach its ramping limit soon unless major investment is made.
  • Loss of frequency protecting inertia as large synchronous plants close, primarily coal and nuclear. As the available inertia diminishes the rate of frequency variation speeds up and so the response needs also increase progressively.
  • Reduction in our ability to keep local voltage within prescribed margins due to the contingent loss from these closures of concurrent reactive (wattless) power capability.
  • High voltage (HV) grid protection has a diminishing ability to ‘see and locate’ faults and monitor the rate of change of frequency from increasing and geographically expanding embedded renewable generation i.e. generation connected to the lower voltage Distribution network rather than the high voltage Grid network. In the UK this has grown following the government’s 2011 capacity mechanism initiative to increase wind back up, which has resulted in the increased use of embedded polluting diesel.
  • Loss of local grid strength (fault level) as large synchronous generation is closed which risks the ability of existing grid protection to properly function/discriminate when system faults occur.
  • Billions are being spent on high capacity high voltage direct current (HVDC) transmission to route wind generation to load. These non-synchronous generators and the HVDC/AC (alternating current) converters have significant operational issues in areas of low fault level as grid system strength decreases. It has been known since 2017 that HVDC/AC voltage sourced converters (VSC) have an issue where system faults relatively nearby significantly increase the risks of loss of stability, frequency control and shutdown. Also with VSCs, rising DC fault currents need AC circuit breakers at both ends to be opened. After the fault has been removed, the HVDC link must be re-started. This is a critical operational point.

In addressing these emerging VSC stability loss concerns, Scottish and Southern Electricity Networks (SSEN) have designed and integrated several levels of ‘bespoke mitigating’ technology to counteract existing low fault levels in the north of Scotland which recur when new nearby offshore and onshore windfarms are becalmed. Others across the UK and Europe may need to follow if the SSEN initiatives prove successful, but it will be technically challenging and very project specific. SSEN are now also to incorporate a giant flywheel at the load end of this £1.1bn HVDC link to seek further system stability (see below).

  • Alternative HVDC/AC Line Commutated Converter (LCC) interfaces, already in wide use, whilst cheaper than VSC, cannot commutate into a dead network so do not support ‘black start’ of a collapsed grid. They also suffer commutation failures from drop or phase shift in AC voltage causing shutdown.
  • After an onshore blackout, AC transmission from large scale offshore wind generation to the grid is not possible from the windfarms alone as they cannot provide the reactive power necessary to re-charge the interconnecting cables. Since the onshore grid is dead it cannot assist. The first procedural action in an area grid blackout is therefore to disconnect all such local offshore generation.
  • The inability of windfarm generation, offshore and onshore, even if not becalmed, to meet grid block loading as loads/districts are re-connected after a blackout means they cannot support grid re-start (e.g. in a ‘black start’). The network has to be re- established using other sources such as gas and coal before windfarms can be re-connected.
  • This means that the lack of large scale, quickly despatchable synchronous generation such as coal and gas in the UK is also now exposing the country to an increasing ‘black start’ problem. Wind and solar renewables cannot re-start us and additional VSC interconnectors to Europe that would commutate into the dead UK network are not yet built. Two of the existing three UK/Europe interconnector designs cannot fulfil this task.
  • The larger the number of HVDC links to an area, the larger the known problem caused by ramping up power supply. If there are several interconnectors needing differing ramping rates huge imbalances can become a problem particularly if unscheduled AC fluctuations occur.
    Stabilising the inverters on large scale solar farms to meet grid voltage and frequency control requirements in remoter areas of low grid strength is another issue. Proven lack of system development modelling in Australia led to a 7 month 50% output curtailment was being imposed by AEMO (Australian Energy Market Operator) on 5 major solar farms in September 2019 following ‘unprecedented grid performance and stability issues’ and fears of uncontrollable oscillations from their inverters if a major outage reduced system strength in their vicinity. In April 2020 AEMO granted only provisional output increases following major modelling efforts but still regard it as a ‘learning curve’. Expensive synchronous compensators are now being installed at many wind and solar farms throughout Australia to help mitigate such issues. AEMO has warned up to twelve other large scale wind and solar projects that there may be delays of up to a year in allowing grid connection. There are also reports of projects being told to delay construction until longer term solutions such as new transmission links are built, which could take up to seven years in some areas. Investors are reportedly becoming more hesitant.

All these issues are becoming more acute as we seek to increase generation from renewables.

How has this happened?
In many countries, certainly in the UK, there is no centralised engineering body responsible for the holistic overview, modelling, determination, and provision of an energy system to best meet the needs of that country, all elements considered, including carbon reduction, and the safeguarding of the security of supply of the national network. In general, of course, outwith power system professionals there is little understanding and appreciation of the real growing challenges at large and the as yet unsolved technical constraints.

UK Carbon Reduction
During the recent UK ESO claimed ‘longest period of coal free operation in Britain’ since the Industrial revolution, the UK imported 10 per cent of its electricity from Europe. However, much of the imported power is coal-generated so during this claimed ‘record setting’ the UK electrical power was not ‘coal free’.

The UK has been working towards a further 6 or 7 interconnectors to mainland Europe which will lead to us increasingly ‘offshoring’ our CO2 as we address ‘Net Zero’ since the carbon data for imported electricity is not counted in UK statistics. This is increasingly referred to in the profession as taking the “least cost pathway to decarbonisation”. It may avoid the UK generators paying a carbon tax on fossil-fuel generated power within the UK, but does little for global emissions of greenhouse gases. It is not just the UK which is pursuing increased reliance on interconnectors.

So what future for the UK as an example?
As we increased unpredictable/ non- despatchable/ and now embedded (distributed) variable renewables our grid is under growing pressure and we are having to play catch up. Yes, renewables have their place but always within a stable grid.

The major grid failure on 9 August 2019 is an illustration of several of the problems described above coalescing. A lightning strike triggered the failure of Hornsea windfarm and Little Barford CCGT power station. As a result, the National Grid lost control of frequency, supplies to 1.15 million customers failed or were ‘load shed’ (disconnected) and around 30 trains tripped on low frequency. Most of these trains had to be reset manually which took hours and caused cancellation/part cancellation of around 590 services. Neither Hornsea nor Barford could meet their Grid Code “ride through” obligations and the Electricity System Operator (ESO) had insufficient reserve available. Consequential tripping of embedded generation on falling frequency was also a major contributor.

In January Ofgem (the government regulator for the electricity and natural gas markets in Great Britain) completed their initial research into the incident and tabled 50 pages of concerns of national importance across the roles and engineering of the ESO, the two generating companies and two distribution companies involved. The report amounts to a relentless and precise identification of myriad technical and procedural failings across the board.

  • Significant ‘lessons learned’ cited by Ofgem include:
    • “The outage highlights the risks and challenges of managing system security and stability in the evolving electricity system as well as the importance of robust industry processes”
    • “The ESO should have been more proactive in understanding and addressing issues with distributed generation and its impact on system security”
    • “Where the ESO is connecting complex power systems to the network, it must be capable of modelling their performance when the network is disturbed. In addition, the process for understanding and ensuring the issues behind fault related outages prior to reconnecting generators should be made more robust.”
    • “Further improvements to the ESO structure and governance framework should be considered in order to meet the challenges of the energy transition”
    • “The ESO should consider and come forward with recommendations to improve …..its holding of reserve, response and system inertia…” (This includes balancing services.)
      Given its serious concerns Ofgem stated they would review the roles of the System Operators in 2020.

Renewables penetration has reached the stage where the ESO are publicly advising that the Grid needs more inertia and reactive power capability to support frequency and voltage stability as a result. As large synchronous stations close, the position worsens. All of the UK’s hugely stabilising nuclear stations, bar Sizewell, will be closed within the next few years. Many of us in the profession have been predicting this inevitability for many years and the ESO similarly would have been aware throughout. The ESO’s declared approach is to now call for this to be actioned and inertia and reactive power to be purchased by the grid users to pay for it. This is a direct consequence of increasing renewables and will further add to the real costs of renewables. But first this capability has to be engineered and made available.

Scottish Power have been talking since October last year of central Scotland likely requiring up to 7 x 300 MVA synchronous compensators with large flywheels to protect their area of the Grid. These would also provide a much needed system fault infeed of 7000 MVA to protect system protection operation: this would address the stability reduction that followed from the closure of Longannet coal-fired station. I know of no other country in the world contemplating this relative scale of deployment. A prototype is at the research stage near Glasgow, so who knows what this will cost, how long it will take and if it will prove to be required widely across the UK? SSEN expect their proposed additional north Scotland flywheel to cost around £25m.

The ESO self termed ‘record’ length of ‘coal free operation’ is so much more complex than it first appears, and as highlighted by Ofgem, increasing the proportion of renewables on the grid raises important issues that require the ESO’s more urgent attention.

David B Watson is a Chartered electrical engineer who was in the energy business for 35 years 30 of which within Foster Wheeler Energy ltd where before retirement he was Manager of Projects of their Glasgow Operations .

David is a Fellow of the Institution of Engineers in Scotland and a Corporate Member of the UK Institution of Engineering and Technology and has been widely published within the UK on power systems issues.

This article is primarily a composite of two of his recent articles for the IET and the Scottish Herald newspaper Agenda column for both of which he has written many times.