**5. Discussion**

In this study, an expanded methodology aims at providing a coherent quantitative assessment of whole districts based on their overall energy storage capacity, load shifting potential, their ability to actively interact with the energy grids and the resulting CO2 emission savings compared to a non-interactive system. As the previously published methodology has been discussed with selected stakeholders, an improvement has been undertaken in order to better adjust the SRI to efficiency standards. In the new version a cut-off for the efficiency of the storage system has been introduced so that large scale, but potentially inefficient storage systems cannot easily substitute smaller, but efficient ones. This is of particular importance in order to avoid that outdated and potentially environmentally harmful technologies are rewarded with a good result. Thus, the efficiency limit, as stated as one of the research questions, has been addressed.

The subsequent enlargement towards a district assessment demonstrates that the methodology can be adapted to bigger and spatially logical entities. The application of the methodology in a theoretical case study shows that the approach is suitable to define relevant indications of the load shifting potential by means of *SRIDist* and *LPDist* as both indicators deliver meaningful results in this context. A weighting factor that has been introduced for the *SRIDist* ensures that districts with different characteristics in terms of size, type and quality can be fairly compared. With this measure, an entity that has a low SRI can be avoided, but high energy demand cannot at the same time within a district be easily compensated with a single entity that has a high SRI but low energy demand. Consequently, the methodology has shown to be meaningfully extended to groups of buildings, thus answering to the second research question.

The *SRIDist* can also function as a benchmark as whole districts can be assessed based on their possibility for further development without relying on measured or monitored data from the energy providers. As outlined in Figures 7 and 8, for example Scenario 1 and Scenario 3 can be compared based on their potential for expansion. Whilst Scenario 3 features twice as many buildings as Scenario 1, the *LPDist* is the same for both scenarios as the load shifting potential is the same across both districts as only half of the buildings in Scenario 3 provide storage capacity. However, from the *SRIDist* it can be clearly seen, that in Scenario 3 there is a considerable higher potential for further load shifting as the *SRIDist* is considerably lower in Scenario 3 than in Scenario 1. This shows that the *SRIDist* can provide a meaningful assessment and comparison of districts that usually greatly di ffer in size, type, as well as quality and number of buildings.

The approach is however solely focused on the load shifting potential and ability of the district to actively interact with the grid. This di fferentiates the methodology from other more interdisciplinary approaches such as those outlined, for example, by Garcia-Ayllon [60] and Sharifi [61] where many other factors such as, e.g., resources and governance, have been taken into account. Also, other schemes such as the BIQ [28] and the HSBS [29] rating, that have been comprehensively compared to the SRI by Apanaviciene et al. [27], deliver a much larger, complex set of indicators that are rather dependent on system and device assessment. Our observation o ffers information targeted on load shifting, thus providing a very focused assessment. The approach outlined in this paper is also not technology specific, but is based on certain qualities, that need to be achieved. This is crucial, as it decouples the methodology from the technology used and makes the approach future proof, as any new technology or device can be similarly represented within this novel approach. The methodology in this paper might therefore serve as one of a series of other quantitative indicators that can be integrated into broader assessments.

In relation to the possible CO2 savings as stated in the last research question, the methodology also provides a workable approach. When considering the systems of energy generation, energy transmission and consumers in relation to the same three systems and adding energy storage, a potential for CO2 savings can be generated. The assumption is that energy which comes from renewable sources cannot be produced on demand, but when it can be stored, it can be subsequently released on demand. It follows that there is the possibility that the magnitude of the LP energy from CO2 neutral but not demand-driven energy sources can be used. This results in a reduction in CO2-related energy and thus an indirect reduction in CO2 emissions from energy production. This definition provides a quantitative assessment of the proportional CO2 savings in relation to the load shifting potential.

Whilst the proposed indicators can be applied for di fferent queries, the following key questions can be answered with the proposed approach: In which area does a high potential for load shifting measures in buildings exist? Which areas already provide enough smart buildings and renewable energy capacity in order to improve the network infrastructure? Is there enough load shifting potential from buildings near, e.g., a wind farm, in order to actively use it as storage? What are the prospective CO2 savings associated with the potential load shifting capacity of the district? For those and similar questions, the proposed initial assessment can provide meaningful results.

There are however also certain limitations to this approach. For one, the assessment of the *LPDist* is not based on actual load profiles and should therefore not be used as a substitute for an exact analysis. Whilst the methodology serves very well for an initial and quick assessment of the load shifting potential within a defined district it is neither intended nor suitable for detailed capacity sizing of energy infrastructure. Care should also be taken related to the choice and extent of the area (system boundaries of the district) under consideration. The *SRIDist* evaluates only the load shifting potential of buildings and does not take any information regarding the respective electrical, thermal or gas grids into account. Thus, any bottlenecks in the network infrastructure are not recognized. This means that just because the buildings are able to move a certain amount of energy it is not necessarily the case

that the networks are also able to accommodate this. It is subsequently evident that the proposed indicators are centered around the assessment of buildings (or multiple buildings) rather than the energy infrastructure.

For further research, a verification of the indicator could be done with a comprehensive dynamic building simulation for the assessment of load shifting capacities, which is based on monitoring data from a wide range of different buildings types. Based on this assessment, the discrepancies of the monitored and subsequently simulated data with the results from the SRI methodology could be calculated. With this, the functions of the model could be calibrated in order to match the actual requirements. In addition, monitoring data derived from buildings varying in typology and energy profiles could be used to assess how well the calculated approximation matches the actual building data. A validation of that kind is planned for the future, once the SRI has to be implemented for Building Regulations purposes. Additionally, a similar indicator that assesses and highlights the free capacities in different network areas or network nodes could be considered. Conversely, one could also estimate for an area which SRIDist\_Max or LPDist\_Max the network infrastructure can endure and thus monitor the development of load shifting potentials and renewable integration in the area in order to be vigilant of potentially critical conditions related to the energy infrastructure.
