Concluding remarks - Modelling Distributed Energy Resources in Energy Service Networks - page 167

8.3 Future work suggestions

By testing the proposed modelling framework in a number of small-scale examples, the
potential of the TCOPF program has been confirmed. However, further insights into
the issues that arise from modelling future energy service networks can be achieved
by strengthening the modelling tool. Enhancing the modelling framework will lead
to more thorough assessments and to the development of other closely related strings
of research. Therefore, future work concerning the modelling of integrated energy
infrastructures with embedded technologies should focus on the following subjects:

Improving input data : Several assumptions from this work have been a result
from data limitations. As a consequence, the network models are a simplified
representation of the physical infrastructures; serving to simulate energy flows at
both medium levels of voltage and pressure, and thus not achieving high levels
of granularity. This is due to the fact that it is difficult to obtain accurate load data
and network topologies that would facilitate a more comprehensive analysis of
the energy systems. Work in this area should focus on developing reliable thermal
( i.e. hot water and space heating) and electric power load modelling profiles for
different sets of consumers. Similarly, profiles on energy driving demands for
improved PHEV representation need to be further elaborated.

●

Incorporating uncertainty into data parameters : The framework proposed uses
a deterministic integrated energy system model which the TCOPF program
resolves and provides an optimal solution. However, in order to gain complexity
and simulate conditions closer to reality, the TCOPF solver should be able to make
decisions on imprecise data, such as load demands, spot market energy prices
and accessible DERs. Consequently, this effort would direct the research towards
a stochastic optimisation model, which in turn would provide probabilistic data
on key variables.

●

Implementing additional DER modelling features : This work showcased the
operation of embedded technologies through linear efficiencies. However, the
inclusion of non-linear efficiencies into the performance of DER technologies
would greatly enhance the description of dispatch patterns these devices would
follow. Hence, a more realistic description of DER operation would allow us to
increase the granularity of the analysis. Furthermore, these actions would poten-
tially identify further economic efficiencies that might be developed in future
power systems.

●

PHEV agent-based modelling : The case scenarios modelled in Chapter 6 are
greatly simplified by assuming that PHEVs are only allowed to charge and
discharge in the nodes where their user resides. However, current aggregation
models depicting PHEV energy demands are insufficient to deal with the data
DNOs will require in order to successfully adopt PHEV load. Addressing mobil-
ity behaviour of users and how their actions will translate into energy demands
within different parts of the network will be an attractive novel field in power sys-
tems, as shown in Chapter 7. Therefore, it can be suggested agent-based models
linking the transport and energy sectors seem to be the most probable approach
to solve this issue.

●