Grid connections are increasingly being recognised as the main barrier to solar PV development, especially at large scales. Solar PV above 3.68kW per phase requires a DNO approval under the G59 commissioning process. There are a number of emerging challenges facing the connection of larger (12kW+) solar PV systems. The main issues emerging in relation to grid connected solar PV are as follows:
1. Volts rise
The DNO’s are legally bound to supply electricity to their customers under ESQCR (Electricity Safety, Quality and Continuity Regulations). One aspect of this is that they must not supply customers at a voltage level outside 230V +/-10% (207-253V) per phase. White goods (fridges, freezers etc.) are designed around this bandwidth across Europe. The DNO’s have to comply with this at both peak and minimum demand. UK electricity demand is lowest in summer often at a time of maximum PV generation. Thus, PV systems can sometimes raise the voltage on the network to which it’s connected.
When assessing the suitability of a connection the DNOs model the system including future (consented) connections to ensure the proposed system does not put the network at risk of falling out of compliance.
Network Reinforcement (i.e. bigger cables and overhead lines with lower impedance to reduce voltage excursions from the nominal value) – this involves more copper/aluminium at significant cost. It is distributed generation (DG) projects, not demand side customers that carry the cost of network reinforcement, although there is some benefit for both sides as the grid is upgraded.
Localised storage and export limiting/constraint. Storage techniques can help smooth out the peaks and troughs, which helps to keep the local network within limits. Limiting can also be useful to provide assurance to the DNO that the system will never pump out more than a predefined power onto the network. The more power injected, the more the voltage tends to rise.
2. Thermal capacity
This is an issue when the power generated is larger than the rating of the cable/overhead, line switchgear and transformers supplying the existing demand based load. Though the voltage rise may keep the network within regulated values, the ratings of the cables and transformers/switchgear may not be rated to transmit the generated electricity in opposing direction to local demand. This is more likely to be an issue where a large stand-alone DG system is planned but is less likely to be a major issue with commercial rooftop PV (where there is greater self-consumption of electricity generated, and the available space for PV is more likely to aligned to the buildings energy requirements and existing grid connection arrangements).
Network Reinforcement. As mentioned above, DG projects carry cost of network reinforcement, not demand side customers. However, as PV is not a 24/7 technology, DNOs could look at PV generation cycle related cable ratings to reduce upgrade costs as the maximum solar generation can only ever be for 16 hours a day from a 24 hour daily cycle. Therefore thermal derating from a 24/7 continuous rating may help reduce connection costs – again power export limiters could be useful here.
3. Fault level impact
Over the last few months fault level impact has become a major emerging issue, and is typically more relevant in urban areas. Fault levels are essentially the levels of current that the components of the network can withstand for short periods until safety devices operate and disconnect the faulty circuit from the grid. The solutions are more challenging as they involve network system upgrades at 33kV and 132kV level. At the time of a fault the DNO examines the fault current impact at sensitive locations on the network from both demand and generation customers on the system. Where, for example, a short occurs, any local generators (e.g. PV systems) will attempt to provide current to support the grid voltage (i.e. fault current), and the collective sum of this fault current from multiple PV systems may exceed rating of DNO (District Network Operator)/TSO (Transmission System Operator) equipment. In many areas the acceptable fault level is expected to be reached or even notionally exceeded with all the DG connections that have been granted permission.
There are some fault current limitation pilots being undertaken by DNOs.
Network reinforcement in form of switchgear and busbar upgrades is another solution.
An investigation into the actual fault current impact of solar PV systems on the network, in particular a more granular review of inverter fault level outputs, and impact on DNO’s and array orientation as this will have real time impact on each system fault current contribution.
4. Network harmonics
The harmonics output by PV inverters can have an impact on the electrical quality on the grid by causing distortions of the voltage waveform. As the installed PV capacity becomes more significant, the collective impact of harmonics starts to become an issue.
As harmonics can cause issues that affect Electricity Safety, Quality and Continuity Regulations (ESQCR) installing harmonic filters can mitigate the problem and it is likely that this measure will be needed. Ultimately, if PV is to be a universal contributor to the nation’s energy demand in every household, public, rural and commercial building, the DNO network will need some degree of harmonic filtering.
5. Real time network information
The DNO SCADA (automated communication system) covers the 33 kV network and above, but the DNOs have limited real time knowledge at 11kV and below. The network planning models are based on ‘worst case’ i.e. all generators running at maximum output. The reality is that it is rare for PV and wind to generate their maximum outputs simultaneously. For example a 15 degree commercial roof will rarely if ever output at rated capacity; and an east facing roof won’t generate the same output at same time as west facing roof. Thus, in reality, on a real time basis, there is probably a significant amount of capacity available that has been allocated a peak output allocation that is rarely achieved. A study by Western Power Distribution (WPD) of the 11kV network in the South Wales area, with a high proportion of domestic PV, demonstrated a smaller impact on the local grid than anticipated due the variance in orientation of arrays from East to West. The different systems generated peak power at different times of the day, so the cumulative impact was found to be less than expected.
Techniques such as ‘timed generation’ and ‘active network management’, where systems are monitored and real-time limits imposed are good examples of implementing real-time, variable export allowances. For example, WPD has recently launched its ‘alternative connection’ service that allows for different types of export control mechanism.
6. Consented capacity
When DNOs appraise an application for connection of a PV (or other DG) scheme they look at the existing and forthcoming connections of which they are aware. This does not necessarily equate to the real picture, but from their standpoint it is the only way. It is likely there is a large amount of capacity allocated for projects that, in reality, will not be built due to changes in government policy and available revenues. The DNOs are now actively attempting to claw back this unused capacity, which could then be reallocated to roof top projects, but it is a slow and delicate task. Currently, DNOs can only withdraw consent if they see demonstrable non-movement on a project.
In this emerging area some potential solutions might include:
A centralised arrangement could be implemented whereby planning approvals, grid approvals and other barriers are managed in a coordinated fashion. Projects that can proceed in line with DECCs solar strategy can then be given preference over projects with grid consents that are not mature, or are detrimental to development of other projects that are aligned with the DECC strategy.
A review of grid consents in comparison to actual capacity installed and registered under the Renewable Obligation scheme (RO). The review would need to be timed to take a snapshot of capacity at the RO closing date for new systems larger than 5MW (31 March 2015). This would allow consideration of available grid capacity to be released after larger projects are downgraded to less than 5MW come April 2015.
An automatic preference for community, national/local authority and rooftop solar projects, or projects with higher levels of self-consumption.
Reserve pots of capacity, ring fenced for roof and community projects, to facilitate market development.
BRE National Solar Centre is working to prepare a full briefing paper on the subject of grid which expand on the topics covered here. For more information contact firstname.lastname@example.org and visit www.bre.co.uk/nsc.