Learnings

Outcomes

The project successfully built a probabilistic simulation algorithm which can model the impact of defined LCT demand profiles on a range of representative LV networks.  A Monte Carlo simulation randomly allocates demand and LCT profiles on a representative network.  Power flow simulations are then processed on the representative network to assess the impact of Heat Pumps and Electric Vehicles.

In terms of ED2 planning and future proofing the network, the representative networks studied showed the LV network approaches its breaking point between the years 2027-2030, assuming they are operated in the existing state without reinforcement.

The study selected Rural Village Overhead, Rural Village Underground and Suburban Street as the representative feeders which are most likely to be affected by the installation of Electric Vehicles and Heat Pumps.  The study provided the following outcomes when considered against Future Energy Scenarios.

* Rural Village OH: For an EV penetration exceeding 25% and by the year 2030 (with reference to FES-Community Renewables and Two Degrees scenarios); For the HPs the network is projected to see upwards of 32% penetration of HPs by the year 2038 (in the FES-Community Renewables) by which time the existing network is likely to approach a breaking point, due to the integration of HPs.

* Rural Village UG: For an EV penetration exceeding 12% and by the year 2027 (with reference to FES-Community Renewables and Two Degrees scenarios); For the HPs the network is projected to see upwards of 32% penetration of HPs by the year 2038 (in the FES-Community Renewables) by which time the existing network is likely to approach a breaking point, due to the integration of HPs
* Suburban Street: For an EV penetration exceeding 25% and by the year 2030 (with reference to FES-Community Renewables and Two Degrees scenarios); For the HPs the network is projected to be safe till the year 2040 (with reference to FES-Community Renewables and Two Degrees scenarios) or for up to 40% of HP penetration on the networks

A penetration of 25% of EVs can be accommodated by the majority of feeders on Northern Powergrid networks with minimal reinforcement costs for the feeders with a low probability of violation.

* On the representative networks studied we see that as EV penetration levels increase, the probability of violation also increases meaning that there are more kilometres at risk of thermal violation and customers at risk of voltage violation. Of the three network types investigated, Rural Village UG network is most vulnerable to integration of EVs and HPs due to both higher customer numbers and different electrical characteristics. Whilst Suburban street was the most robust.
* On rural village networks there are select cables on the network that are more vulnerable to overload, particularly the cables closer to the substation transformer. These cables although not of considerable lengths have high probability of experiencing a significant overload. Furthermore, in some Rural networks on NPg there are possible clusters of EV integration (within the possible diversity of EV locations) that are susceptible to see overloads due to higher EV charger size.
* Due to a more dense layout of Suburban street networks the line that are overloaded are considerably short there are several lines on the network that are simultaneously at risk  overload albeit with a relatively lower probability of violation and therefore the numerical value of the total kilometres at risk (scaled results in table) do not fully reflect the severity.

* From the variety of battery sizes studied in the simulation , it is EA Technology’s view that in addition to the electrical parameters the customer demographic and material performance of the car/battery manufacturer has a significant contribution to the health of the network in terms of risk of violation

*A penetration of up to 40% of HPs can be accommodated safely on Northern Powergrid suburban networks with minimum reinforcement costs for the feeders with a very low probability of violation. However, on Rural networks are more vulnerable to integration of HPs wherein a penetration of 25-30% of HPs could lead to instances of a LV planning limit violation.

* The simulation results indicated that the Cold Load Pickup event does not have a substantial impact on the network and the violations observed on the networks are mainly down to the additional load introduced by the integration of HPs. During winter months, to maintain a minimal comfortable temperature, it is likely that a HP system will be operating close to its maximum utilisation during periods of home occupancy resulting in a low diversity of HP loads. While this puts a strain on the network, due to this high base utilisation, however, the HPs were not a significant source of additional strain during an outage cold load pick-up event.

* In some rural networks the probability of a violation an outage is greater than the probability of violation with an hour outage at 5:30 pm. Therefore, during periods of significant cold weather, when network load is traditionally highest, outage events are unlikely to have a strong impact on network after restoration and in someinstances alleviate the risk of violation during peak loading.

* It is possible that during milder weather such as winter average it is more likely outage events will have an impact on heat pump Cold Load Pickup. This is because when weather is milder (and network load is traditionally lower), a heat pump will have a lower level of utilisation during normal practice and therefore a sudden loss of diversity due to a CLPU could be more significant.

* A comparison between the simulation results and the corresponding values in IMP/001/911 Northern Powergrid Code of Practice (CoP) for the Economic Development of the LV System showed that the ADMD and maximum network demand values obtained were significantly different. The values in the CoP were considerably higher than the results obtained by the engineering analysis using real data. Although the framework calculation was not aimed to differentiate or suppressed the existing values/guidelines using ACE49 and ADMD (for nth customer) methodology used by Northern Powergrid, it is clear that with changing nature of customer demand a detailed study of the underlying assumptions used for existing network design methodologies both ADMD and ACE49 is pertinent to overcome the challenges posed to LV planning and new customer connections.

Lessons Learnt

N/A