Boštjan Blažič

1. Introduction
In the field of electricity supply, we are witnessing big changes, which are the result of the proliferation of renewable energy sources (RES), the electrification of transport (electric vehicles – EVs), the lowering costs of electricity storage and the increase of the share of heat pumps. These changes mostly affect the operation of distribution networks that were not designed for such operating conditions. Namely, an uncontrolled and uncoordinated operation of new devices may lead to problems with the operation of a network, especially due to increased power flows and voltages outside of the defined limits.
For a reliable and efficient operation of the power system and for avoiding over-sizing of the network, a higher power system flexibility is required. In addition to traditional sources (large power plants), flexibility in the electric system can be provided also by modern concepts of distribution network operation (e.g. coordinated voltage control) as well as by active network users, this is the users that are able to adjust their consumption or generation to network conditions or to market requirements. Measures should be developed to enable the active users to participate in network control in order to resolve as many critical operational states as possible.
Using the potential of active users in the power system requires analysing a multi-service approach, where several types of stakeholders of the market are involved: regulated operators that deliver energy to end users (TSOs and DSOs), aggregators and balance responsible parties that respond to power system needs, and, in the end, industrial consumers and households where some of them became prosumers since owning their own electricity generation (like PV units). When it comes to the distribution network, two main challenges must be addressed:

• The market entities controlling active users (i.e. aggregators) are usually not aware of the operating conditions of the distribution grid: distribution operators face physical constraints (power flow, voltage levels) which impose limits to users’ operation (e.g. EV charging). Such constraints must be taken into account when active users envisage to provide system services to meet the needs of distribution network operators (like congestion management, voltage control or reactive power provision).
• An efficient distribution network management requires that the distribution system operator is able to define the required power from multiple and distributed active users in view of serving the distribution network at appropriate costs.

Addressing the high variability of consumption and generation, as well as adherence to modern network control methods, which also include services, clearly requires an appropriate system architecture and a simulation platform that enables the evaluation of different solutions and forms the basis for network planning and market design.

2. Smart-grid system architecture
Services for different stakeholders of the electricity market can be provided by active users, which can include EVs, electricity storage, flexible demand, RES… EVs are clearly one of the ‘flexible users’ with a high potential for services provision by means of shifting charging times, however, other sources of flexibility can share the same system. A possible smart-grid system architecture that enables the provision of services is illustrated below.

Figure 1: Smart-grid system architecture for provision of services

The main components and requirements of the architecture are:

• The aggregator aggregates the active users’ flexibility. Clearly, the flexibility of a user is mainly due to its ability to shift consumption in time. Therefore, the process of flexibility aggregation is a process of scheduling user consumption (e.g. EV charging) as a function of time.
• The DSO has to be able to calculate the distribution network constraints. As services have a duration in time, the constraints have to be computed in real time and forecasted for a short-term future (day ahead, 4-hours ahead…). The constraints are communicated to aggregators or other operators of active users.
• The aggregator has to be aware of the distribution network constraints that may limit the provision of services.
• A DSO or a TSO can use the flexibility of active users as a service supporting grid operation.
• A market operator can perform the coupling of service offers and service requests (essentially an optimisation problem). However, services can be provided also on a direct contract basis (e.g. TSO - Aggregator).

Basically, two types of operation are associated with the power-network services system:

• On-line (in-field) operation of the system in real time.
• Off-line operation of the system – simulations of power system operation used for planning purposes.

2.1 On-line network operation (real-time operation)
According to the proposed architecture of services provision, one of the main tasks, in terms of development, is the enhancement of the DSO observability, which includes:

• A sufficient number of measurements in the distribution grid.
• A robust state estimation algorithm for assessment of network operating conditions in all network feeders and nodes.
• A forecast algorithm for the assessment of short-term future network operating conditions.
• A calculation of the distribution-network transfer-capacity, which calculates the network constraints (in real-time and for the short-term future), which may constrain the access to flexibilities.
• A calculation of the required distribution-network services, which calculates the services that will be needed by the DSO to ensure a safe network operation.

On the aggregator’s side, two additional functionalities are needed:

• A distribution network constraints module, which enables the aggregator to be aware of the constraints which may hamper the use of its assets.
• A services module which enables the provision of services that can be offered to power system entities.

A network observability tool is essentially a real-time tool operating on-line in a distribution network and involves three critical functionalities:

• The distribution-network state estimation
• The forecast of load consumption and voltages
• The distribution-network state forecast

The required input data for the observability tool are:

• The regularly updated network topology
• The network elements basic data (transformers, power lines, storage)
• The real-time voltage measurements in at least one point of the network
• The real-time power measurements in multiple points (load, storage, DG)
• The historical consumption/generation data
• The weather forecast

The state-estimation and state-forecast tools calculate the voltages at all the network nodes and power flows in all feeders, all in real-time and for the short-term future. The data is fed into two modules:

• The distribution-network transfer-capacity module: the module calculates the actual (real-time) and short-term future network constraints, which is based on the maximum feeder currents and allowable voltage limits. The output of the module is the maximum infeed or consumption of an active user at a particular network location.
• The grid services module: the module calculates the required actual (real-time) and short-term future power from the storage units in order to provide the expected network support.

The network observability tool can be a standalone tool or a part of a Distribution Management System (DMS). In the case of a standalone module, it is programmed using an industrial grade controller, and includes state estimator and forecasting algorithms. It receives the network data from the DSO’s SCADA/DMS system, covering the network area of interest, and calculates the network margins and required services. The calculated data are available for aggregators through an API module as a standardised interface. In case of an integration into a DMS system, the tool uses the DMS modules, especially the state-estimator and the forecasting algorithm. The calculated network margins and/or required services are available through the same API interface.

2.2 Network planning tool
The basic structure of a planning tool is shown in the figure below. An important focus should be on proper modelling of network users, that is, consumers and sources. Each user can be passive or active, so the user model must enable both modes of operation. An active user is one who enables the variation of their consumption or generation according to network operating conditions or to market requirements. Consumers and sources in the network are described by their diagrams of power consumption or generation, which are given in the form of time series (power as a function of time). The passive user load diagrams are based on historical measurement data, and for active users these diagrams change according to the desired management goals. The main groups of users are:

• classic household consumption,
• heat pumps,
• electric vehicles,
• distributed generation and
• electricity storage.

For the calculation of power flows operating scenarios can be defined. In each scenario different variables are defined, for example: the location and the number of RES and electric vehicles, season and type of day (working day, weekend ...), method of voltage control (at one point, coordinated ...), proportion of active users participating when managing a network by adjusting power consumption, etc.
The power flow calculation can be based on a sequential time calculation or on the Monte Carlo approach, which takes into account the high variability of consumption and generation and the randomness of locations of future users.
The network-planning tool should enable the comparison of different solutions (e.g. grid reinforcement vs active user services) based on the defined KPIs (costs, hosting capacity, losses…) and forms the basis for network planning.

Figure 2: A basic structure of a network planning tool

3. Conclusions
An advanced and flexible power system will require substantial changes in the operation and planning of distribution networks. In terms of network operation, distribution network observability and controllability will have to be substantially enhanced, and, in terms of planning, advanced smart grid solutions will have to be included in the process. Moreover, much more collaboration among different system entities will be required, otherwise the integration of RES and EVs may turn out to be a pricy transition.