**Preface to "Smart Energy, Plasma and Nuclear Systems"**

This Special Issue on the topics of smart energy, plasma and nuclear systems contains a collection of five extended papers from the SEGE (International Conference on Smart Energy Grid Engineering) 2019 and SPAN (Symposium on Plasma And Nuclear Systems) 2019. The SEGE conference aims to provide an opportunity to discuss various engineering challenges of smart energy grid design and operation by focusing on advanced methods and practices for designing different components and integrating these within the grid. It also provides a forum for researchers from academia and professionals from industry—in addition to government regulators—to tackle these challenges and discuss and exchange knowledge and best practices about the design and implementation of smart energy grids. The Symposium on Plasma and Nuclear Systems (SPANS) is connected with SEGE. SPANS provides a forum for researchers from academia and industry to present and discuss the latest research innovations in nuclear and plasma systems. SPANS will provide attendees with state-of-the-art research and technologies, while engaing in active discussions with industry. Additionally, it will provide industry with opportunities to promote their products and business cases. Attendees from regulators and standards will engage in fruitful discussions on how R&D is linked with regulations and standards. The 2019 edition of SEGE was held in Oshawa, Canada, and attracted a total of 112 regular paper submissions, spanning numerous active and emerging topic areas. The conference program committee selected 60 papers to be presented at the conference and published in the conference proceedings. The five extended papers for this special issue were selected from among all the accepted papers by the Special Issue Guest Editor Dr. Hossam A. Babbar, based on the relevance to the journal and the reviews of the conference version of the papers. The authors were asked to revise the conference paper for journal publication and in accordance with the customary practice of adding 30% new material. The revised papers, again, went through the normal journal-style review process and are now finally presented to readers in this Special Issue. We greatly appreciate the willingness of the authors in helping to organize this Special Issue.

> **Hossam A. Gabbar** *Editor*

## *Article* **Energy Analysis for the Connection of the Nuclear Reactor DEMO to the European Electrical Grid**

#### **Sergio Ciattaglia 1, Maria Carmen Falvo 2,\*, Alessandro Lampasi <sup>3</sup> and Matteo Proietti Cosimi <sup>2</sup>**


Received: 31 March 2020; Accepted: 22 April 2020; Published: 1 May 2020

**Abstract:** Towards the middle of the current century, the DEMOnstration power plant, DEMO, will start operating as the first nuclear fusion reactor capable of supplying its own loads and of providing electrical power to the European electrical grid. The presence of such a unique and peculiar facility in the European transmission system involves many issues that have to be faced in the project phase. This work represents the first study linking the operation of the nuclear fusion power plant DEMO to the actual requirements for its correct functioning as a facility connected to the power systems. In order to build this link, the present work reports the analysis of the requirements that this unconventional power-generating facility should fulfill for the proper connection and operation in the European electrical grid. Through this analysis, the study reaches its main objectives, which are the definition of the limitations of the current design choices in terms of power-generating capability and the preliminary evaluation of advantages and disadvantages that the possible configurations for the connection of the facility to the European electrical grid can have. In reference to the second objective, the work makes possible a first attempt at defining the features of the point of connection to the European grid, whose knowledge will be useful in the future, for the choice of the real construction site.

**Keywords:** nuclear fusion; tokamak; generation power plant; power system; electrical transmission grid

#### **1. Introduction**

The European roadmap to fusion energy, summarized in Figure 1, includes the DEMOnstration power plant (generally identified as DEMO), which represents the first fusion reactor designed to supply electrical power to the electrical grid to which it will be connected [1]. With the start of its operation, currently set for the middle of the 21st century, DEMO could be a revolution in the world of nuclear fusion power and in general in the world of power generation.

The first and fundamental step of the roadmap however is ITER, a research tokamak project currently under construction in France that is foreseen to be operative in around five years. ITER is not designed to generate electrical power, but it is designed to achieve five goals that are essential for the continuation of the research in this field [2]:


• It will demonstrate the safety characteristics of a fusion device.

DEMO will largely build on the ITER experience; indeed, its construction will start after several years of ITER operation. Now, the design of DEMO is based on five main objectives [3]:


Considering the achievements that both the facilities should reach, the actual characteristic that distinguishes DEMO from ITER is the size and consequently the possibility of achieving higher values of fusion gain factor and longer fusion time. The fusion gain factor is defined as the thermal power produced inside the reactor during the fusion reaction divided by the thermal power delivered to the plasma during the operation. In particular, DEMO being bigger than its predecessor, it is designed to reach a fusion gain factor between 10 and 50 [1], while this value for ITER is foreseen to be around 10 [1]. One of the aims of current studies on this topic is to understand if this gain factor is high enough to allow a feasible supply of electrical energy to the grid, also because the electrical power systems of tokamaks like ITER or DEMO are larger and more complex than those of nuclear-fission power plants.

As for every tokamak, the main limitation for DEMO is the impossibility of maintaining the fusion reactions for an indefinite time [4]. This is an intrinsic characteristic of tokamaks, and it is related to the need of charging and discharging the central solenoid (CS) system that generates and confines the plasma current. This essentially means that the operation of DEMO is variable, and in particular, it is divided in several phases that will be presented in following section.

**Figure 1.** Summary of the EUROfusion Roadmap to fusion energy.

Now, the researchers are trying to understand how and how much the non-generation time can be reduced, even if the generation time is already foreseen to be at least one order of magnitude greater than the non-generation time. From the structural point of view, the need to optimize this characteristic led to the identification of two possible alternative configurations. The first one involves the direct coupling between the Primary Heat Transfer System (PHTS, the system that extracts the thermal power from the reactor walls) and the Power Conversion System (PCS), which reflects the variability of the thermal output into the electrical power output. The second one involves the indirect coupling between the PHTS and the PCS, which allows to decouple the variable thermal output from the electrical power output, with the interposition of an Intermediate Heat Transfer System (IHTS) with an Energy Storage System (ESS) based on molten salts [3], allowing a constant electrical power output at the generator level.

The article includes five sections. Section 2 defines the main characteristics of DEMO, and it introduces its operational phases. Section 3 focuses on the electrical generator, defining the limitations of the possible coupling configurations, through the analysis of the European Network of Transmission System Operators (ENTSO-E) requirements for generators. Section 4 presents three possible connection solutions for the DEMO facility, considering both the demand and the generation, providing an overview of the advantages and disadvantages of each solution. In Section 5, starting from the input and output power profiles, some features of the point of connection to the grid are evaluated. Section 6 resumes the conclusions of the study.

#### **2. DEMO Features and Operational Phases**

DEMO is foreseen to generate a fusion thermal power inside its reactor that has been evaluated to be in the order of 2 GWth [5]. Now, two solutions are under study for the thermal power extraction and accordingly for the PHTS. The first solution exploits a mature technology, which is the water cooling, also used in fission nuclear plants for its simplicity and reliability. The second solution instead foresees the use of helium for the PHTS, which seems to be promising for future applications but is still a relatively new technology. Conventionally, the first solution is identified as Water Cooled Lithium Lead (WCLL), where lithium lead refers to the technology adopted inside the reactor wall, and the second one as Helium Cooled Pebble Bed (HCPB) [6].

In case of direct coupling between the PHTS and the PCS, both the WCLL and the HCPB configurations provide for the implementation of a Rankine cycle for the thermal power conversion. This means that in case of HCPB, the helium cools the reactor (PHTS), and then, it exchanges the extracted power with the cycle working fluid, namely water. Moreover, in case of indirect coupling, the working cycle is a Rankine cycle, but in this case, the decoupling between the PCS and the PHTS makes the overall operation independent of the type of fluid circulating in the PHTS. From now on, more efforts are being focused on the WCLL configuration and so more data are available. This study will mostly refer to this solution. In any case, several results and procedures can be applied both to the WCLL and to the HCPB configurations.

For the limits of the tokamak technology, the thermal power is generated only during a portion of the operation time, which is generally identified as the "Burn Flat-Top" phase. A smaller power can be generated in the other phases for thermal inertia of the materials, nuclear reactions and other phenomena but always, as a consequence of the Burn Flat-Top. In the present configuration, this phase should last around 7200 s (2 h). Between each burn phase, other phases are conventionally identified during which no relevant thermal power is extracted. These phases are resumed in Table 1, with their respective duration.

For completeness, a brief description of the phases is reported. During CS pre-magnetization, the CS, that is, the core of the magnet system, is energized in order to be able to generate a power pulse strong enough to generate the plasma ignition that is represented by the Breakdown phase. Once the plasma has been generated inside the reactor, it must be heated up to the temperatures needed to have a sustainable rate of fusion reactions, and this is done during the Plasma Ramp-Up and Heating Flat-Top phases, mainly using the additional heating (AH) systems. Now, the most likely solution allows for the use of three technologies for the AH: Electron Cyclotron Resonance Heating (ECRH), Ion Cyclotron Resonance Heating (ICRH) and Neutral Beam Injection (NBI) [7]. When the temperature reaches the order of 107 K, the conditions inside the reactor allow to have self-sustained fusion reactions happening, and this identifies the Burn Flat-Top phase. During this phase, the power demand from

the magnet system and the AH is minimum, while the thermal power generated is maximum. At the end of the burn phase, the reactor has to be brought back to the initial conditions avoiding shocks in the plasma, and this is done during the Plasma Ramp-Down and the Dwell time.


**Table 1.** Plasma phases.

The whole operation of DEMO, in terms of input and output power of the facility, can be described referring to the phases resumed in Table 1.

For what concerns the power needed to operate the facility, two types of loads have to be considered: the steady-state loads and the pulsed loads. The first ones require a steady-state 50 Hz voltage input, i.e., auxiliaries, cryogenics, pumps or compressors for the PHTS, etc. The pulsed loads, i.e., magnet system and AH devices, have to be supplied with variable voltage waveforms, and so, they will be provided with complex power conversion systems that are currently under study. The steady state and pulsed loads are supplied by dedicated substations and distribution systems [5], respectively, the Steady State Electrical Power System (SS EPS) and the Pulsed Power Electrical Power System (PP EPS). By convention, the PP EPS is identified with its two subsystems, the Coil Power Supply Pulsed EPS (CPSP EPS) and the Heating Power Supply Pulsed EPS (HPSP EPS), which supply the magnets system and the AH devices, respectively. From the generation point of view, instead, the generator subsystem is here conventionally referred to as Electrical Generator EPS (EG EPS). The connection node and the High Voltage/Medium Voltage (HV/MV) transformation sub-station are defined as High Voltage Switchyard EPS (HVS EPS).

In order to have a preliminary overview of the facility's demand, Figure 2 presents a qualitative profile of the input active power, derived from the data available in the EUROfusion private database.

**Figure 2.** Qualitative profile of DEMOnstration power plant (DEMO) active power demand during one cycle.

Regarding the power generation, of course, the direct and indirect coupling cases have to be separately considered to evaluate the output power profiles.

Despite being the simplest solution form the constructive point of view, the direct coupling between the PHTS and the PCS leads to several thermomechanical and electrical concerns. Regarding the thermomechanical aspects, several studies, not public but available for the authors as researchers involved in DEMO project on the Eurofusion database, assess the impossibility of operating the turbine with a completely direct coupling with the reactor. Firstly, the thermomechanical stresses due to the abrupt changes in the steam mass flow rate would be unsustainable for the turbine, leading to premature failures of the turbine itself. To limit these cyclical stresses, the maximum steam mass flow rate has to be reached with a ramp. In particular, the nominal power of the turbine has to be reached with an increase of 10% of the nominal power per minute. Another limit concerns the minimum power at which the turbine can be operated, that is, the minimum steam flow rate that can be supplied. In this case, the problem affects both the thermomechanical and the electrical aspects. Indeed, from the thermomechanical point of view, the turbine would suffer from the cyclical start and stop procedures, while from the electrical point of view the generator cannot lose the synchronism with the grid, so it has to be kept spinning. This means that the turbine has to be supplied with the proper mass flow rate of steam also during the non-generation time. In particular, the minimum power of the turbine has been set to 10% of the nominal power. During the reactor non-generation time, the solution currently under study allows for the implementation of a small electrically heated molten-salts loop, designed to provide the proper mass flow rate of steam to the turbine.

Referring to the latest studies, not public but available for the authors as researchers involved in DEMO project on the Eurofusion database, the nominal power of the steam turbine has been evaluated to be around 790 MW. Considering a 5% value for the losses due to the coupling between the turbine and the synchronous generator, the nominal value for the electrical power output can be estimated to be about 750 MW. Considering this nominal value and the limits previously reported in case of direct coupling between the PHTS and the PCS, the resulting active power profile is shown in Figure 3.

**Figure 3.** Electrical Generator (EG) active power output profile in case of direct coupling between the Primary Heat Transfer System (PHTS) and the Power Conversion System (PCS).

For the indirect coupling configuration, the situation is more complex from the constructive point of view, due to the IHTS and the molten-salts ESS, but the management of the turbine and of the generator is simpler. Indeed, in this case the steam flow rate supplied to the turbine can be maintained practically constant during the whole operation of the facility, like in a conventional power plant, minimizing both the thermomechanical and the electrical stresses.

Referring to the latest studies available on the EUROfusion private database, the nominal power of the turbine in case of indirect coupling configuration is around 675 MW. Considering the same value used in the previous case for the efficiency of the turbine-synchronous generator coupling (0.95), the resulting active power output at the EG level is around 640 MW. The resulting profile is reported in Figure 4.

**Figure 4.** EG active power output profile in case of indirect coupling between the PHTS and the PCS.

#### **3. Connection of DEMO EG to the European Electrical Grid**

Being the first fusion reactor able to deliver electrical power to the external grid, DEMO has to face new issues with respect to its predecessors. Of course, as it has been mentioned before, DEMO is not a conventional power plant. However, since it uses a conventional thermodynamic cycle for the conversion and is connected to the grid through a synchronous generator, now it has to be considered as a conventional power plant from the regulatory point of view.

In reference to the current European power systems regulation, as a synchronous generator connected to the transmission grid, for a proper operation, DEMO has to fulfil the requirements set by the ENTSO-E. Specifically, these requirements are defined in the ENTSO-E Regulation of 14 April 2016 "Establishing a network code on requirements for grid connection of generators" (commonly said RfG network code) [8].

In what concerns synchronous generators, which in the code are referred to as synchronous Power Generating Modules (PGMs), the RfG distinguishes four categories, according to the voltage level of the connection and to power rating of the facility (identified as maximum capacity inside the regulation). The categories reported in the RfG are the following:


Therefore, for the specific case of DEMO, the requirements for Type D synchronous PGMs, namely those with a power rating higher than 75 MW, have to be considered. The requirements for Type D PGMs deal with:


Since the research is still in a preliminary design phase, there is not enough information to analyze all the aspects of the RfG. Therefore, for the purpose of this study, we will refer only to the main aspects, i.e., those that can have an actual impact on the current design choices. In particular, we will consider the items in the list above from 1 to 7 and 11. The items from 1 to 6 deal with power generation and control, while items 7 and 11, respectively, deal with the capability of restoring the system after a shutdown of the grid and with the reactive power capability of the synchronous generator.

#### *3.1. Constant Power Output Requirement*

The frequency stability requirement defines the frequency ranges, and respective time intervals, for which the PGM shall be able to remain connected and operate in the grid. Moreover, for item 6 in the list above, in this range, the PGM shall be capable of maintaining a constant output at its target active power value. While in case of indirect coupling this requirement is fulfilled (Figure 4), it represents the strongest practical limitation to the choice of a direct coupling configuration, due to the variability of the EG power output. Referring to the profile of the active power output reported in Figure 3, considering the large variation of the output power (from 75 to 750 MW) and the fast variation in time, no compensation could be possible on the electrical side. Therefore, if the direct coupling configuration has to be adopted, the only way to fulfill the constant output requirement is to act on the mechanical power and so on the thermal power provided to the turbine or to act outside the main turbine-generator (TG) group. Therefore, the constant output power can be achieved in two main ways: One way is to make the TG work at the same active power set-point during all the phases; the other way is to couple the TG with an auxiliary generation set, even based on a totally different technology.

The first way can be seen as a category of solutions, all based on feeding a constant mass flow rate of steam to the steam turbine. Feeding always the same mass flow rate to the steam turbine essentially means to have an auxiliary steam generator. Actually, in the direct coupling PHTS-PCS scheme, one auxiliary steam generator is already foreseen, but it only provides the steam needed to operate the TG at the 10% of its nominal power (as mentioned in the previous section). Coherently with the volumes involved, the ESS loop could be foreseen with a higher rating, in order to guarantee up to 100% of the nominal mass flow rate of steam to the TG. Since the ESS loop is currently foreseen to be fed by an electric heater, it has to be considered that the net output power of the facility will be lower. Moreover, the convenience of the direct coupling approach, which is mainly the simplicity of construction with respect to the indirect coupling one, could be no longer that evident. Nonetheless, this solution could be even more complex than the pure indirect cycle.

The second solution is even more complex and expensive with respect to the first one, since it requires the implementation of a secondary generation set able to compensate the generation during the non-production time of the reactor. Since the power rating of the auxiliary generation set should be comparable with that of the first one, this kind of solution appears to be unlikely to be implemented. At the connection with the external grid, the sum of the two generation sets would appear as a good approximation of a conventional Type D PGM, but actually, it would have an installed maximum power capacity which is around two times the nominal one.

#### *3.2. Power Control Capability Requirements*

The Limited Frequency Sensitive Mode—Over/Under frequency (LFSM-O/U) is an operation policy, activated by the Transmission System Operator (TSO) when the grid is an emergency state of over/under frequency and needs a fast decrease/increase of active power generation [9]. The Frequency Sensitive Mode (FSM) instead represents the ordinary operating mode of a PGM, in which the power output changes in response to a change in the frequency of the system, in such a way that it supports the recovery of the target frequency [10]. Therefore, all the points from 2 to 5 deal with the power control capability of the PGM. In particular, the most restrictive value defined in the RfG is 10% (so ΔPmax = 0.1 Pmax) power control capability, which is provided for the FSM operating mode. It is important to stress out that the provision of the power control capability is limited by the minimum regulating level and by the maximum capacity of the PGM. In addition, the ambient conditions and the limitations on the operation near maximum capacity at low frequency of the PGM have to be considered.

Regarding the indirect cycle configuration, while the stable and continuous operation with a fixed active power set point should be guaranteed thanks to the IHTS, evaluations on the molten salts ESS-water coupling should be performed, in order to define the potential of the system in terms of rate of change of active power.

In case of direct coupling, moving from the profile in Figure 3, the continuous and dashed red lines in Figure 5 represent the profile that the generation should be able to maintain in order to operate in the European Network as a conventional Type D synchronous PGM. In particular, the continuous line represents the maximum capacity operation that by definition the PGM should be capable of providing continuously during the ordinary operation in the grid. The dashed line instead defines the active power lower limit for the operation in case of FSM, with the power variation set to the maximum value of 10% of the maximum capacity as a conservative solution, since its actual value depends on the agreement carried out between the facility owner and the relevant TSO.

**Figure 5.** Active power output adapted to the European Network of Transmission System Operators (ENTSO-E) requirements.

Concerning the red dashed line, the capability of lowering the active power output, since the fusion thermal power has to be extracted in any case from the reactor (referring to generation time), it is sufficient to use suitable bypass valves upstream from the steam turbine. Of course, the solution must be compliant with the capacity of the condenser downstream the turbine that should be able to elaborate additional flows of steam and not only the double-phase flow coming from the turbine itself. For the direct coupling configuration, this condition should be already fulfilled since during the power ramps a variable portion of steam bypasses the steam turbine, so there could be no need for adjustments in these terms.

#### *3.3. System Restoration Requirements*

Concerning the system restoration, the relevant TSO may require the black start capability. A PGM with the black start capability shall be able to restart from a shutdown without the external grid electrical energy supply. This kind of service may be possible in case of indirect coupling, depending on the size of the reserve and on the duration of the shutdown, but it is not compliant with the operation in case of direct coupling. Indeed, in case of a network shutdown, the reactor is likely to be shut down in turn, since its operation and security strongly depend on the network supply. Moreover, for the same reason, the reactor cannot be restarted without the external grid, leading to the impossibility of a black start of the PGM.

#### *3.4. Specific Requirements for Type D Synchronous PGMs*

The operation of a Type D synchronous PGM requires additional specifications, mainly in terms of voltage stability and so about reactive power capability at and below the maximum capacity. From these requirements, we can extrapolate the maximum value for the Q/Pmax ratio that a Type D synchronous PGM connected to the European electrical grid should be able to provide. This means that we can evaluate a preliminary power rating, in terms of apparent power, of the synchronous machine that has to be coupled with the steam turbine. Since the maximum value set in the RfG for this ratio is 0.65, which means cosφ = 0.84 (a common value for machines of this rating that can be found also in literature [11]), we will have that:


#### **4. Layout Options for HV Switchyard Including the Generator**

The results reported in this section take as reference the regulation on the connection of new facilities provided by the Italian TSO, Terna [12], considering that the Italian policies are based on the ENTSO-E prescriptions, so this approach is more conservative with respect to an approach based on the European requirements.

Relying on the guidelines provided by Terna, the connection of new facilities to the national grid must be planned in accordance with a proper procedure that is outlined in Figure 6.

**Figure 6.** Procedure for the connection of generators and consumers to the European grid.

In particular, considering that the project is still in a preliminary design phase, the first three steps reported in Figure 6 will be considered in the analysis and so those that deal with the connection planning and not with the executive design, which are:


For the definition of the voltage level and of the connection scheme, the technical document [12] provides a table summarizing standard solutions, partially reported in Table 2.



The in-and-out connection is performed with the implementation of a new station on an existing transmission line. This means that the new station is supplied by two different transmission lines, coming from two different nodes. Therefore, the facility can ideally operate even when one of the lines is out of service. The radial connection is similar to the in-and-out one, but the connection starts from an existing station of the transmission line. This solution is generally adopted when the distance between the existing station and the facility is lower than 10 km.

Keeping in mind this simplified procedure to evaluate the connection characteristics, three possible configurations have been considered for the connection of DEMO to the European transmission grid:


In the following subsections these solutions will be analyzed starting from power profiles elaborated in the DIgSILENT PowerFactory simulation environment [13], moving from the data available on the EUROfusion database in terms of production and consumption of the facility. Concerning the demand, the data refers to the assisted breakdown scenario, which allows for the use of EHCR during the breakdown phase to assist the magnets system, lowering the power peak required for the plasma ignition.

#### *4.1. Single POD*

The single POD configuration is the one that requires the simplest implementation. In case of single POD, DEMO would fall in the category of power generating/demanding facilities. This means that the value of power to be considered in the evaluation of the connection scheme is the highest between the generation maximum power and the demand maximum power [12].

In Figure 7, the preliminary active power profiles are presented, in case of direct and indirect coupling.

**Figure 7.** Power profiles at the HV node in case of single Point of Delivery (POD) configuration.

As shown in Figure 7, in case of direct coupling, the maximum power absorbed is around 500 MW, while the maximum power injected is around 360 MW. The maximum power is reached in "demand mode", and its value is higher than 100 MW. This means (considering Table 2) that the connection shall be performed at the highest available voltage level with the implementation of an in-and-out single bus-bar scheme with bypass. In case of indirect coupling instead, we can see that there is a peak of injected power of around 650 MW, which is higher than the peak of absorbed power, so the connection has to be performed at the highest available voltage level with the implementation of an in-and-out double bus-bar scheme (Figure 8). Of course, also the radial configuration can be adopted if allowed by the distance of the station.

**Figure 8.** In-and-out double bus-bar connection scheme in case of single POD.

With the implementation of a single POD, the node cannot be seen from the external grid point of view as a proper generation node. This means that the current legislation on the connection of generators (RfG code) to the grid cannot be applied. Moreover, this configuration is as easy to implement as it is dangerous, both for the operation of the facility and of the network. In fact, the high-power spikes absorbed by the converters to feed the PP EPS loads would jeopardize the functioning of the generator, since it would be the nearest source of power. This implies high transient electromechanical torque applied to the shaft of the TG, whose integrity could be seriously compromised, both for the magnitude and for the cyclicality of the resistive torque. To avoid this situation, an electrical ESS could be implemented upstream the PP EPS, to limit the power derivatives.

#### *4.2. Double POD with EG-Dedicated Node*

It is clear that the generator should be decoupled as much as possible from the PP EPS supply. One way to do so is by implementing a double POD solution, with one POD dedicated to the EG EPS and the second one for the SS EPS and PP EPS. Figure 9 shows the active power profile at the POD for the SS EPS plus PP EPS, and Figure 10 shows the generation profile of the EG EPS.

**Figure 9.** Power profile at the Steady State Electrical Power System (SS EPS) + Electrical Power System (PP EPS) HV node in case of double POD configuration.

**Figure 10.** Power output of the generator in case of direct and indirect coupling.

For convention, the power supplied by the external grid is represented as positive, while the power injected into the grid is negative. However, in Figure 10 the generator output power profile is defined from the generator point of view, so it is represented as positive. Of course, from the external grid point of view, that power is negative, since it is injected into the grid itself.

Regarding the SS EPS plus PP EPS POD, since the maximum power required from the grid is around 700 MW, the connection shall be performed at the highest voltage level available, with the implementation of an in-and-out single bus-bar scheme with bypass (see Table 2). Concerning the EG dedicated POD instead, the maximum power injected is 750 MW in case of direct coupling and 640 MW in case of indirect coupling. Therefore, independently from the coupling configuration, the connection shall be performed at the highest voltage level available, with the implementation of an in-and-out double bus-bar scheme.

Figure 11 reports the in-and-out connection schemes in case of double POD configuration with EG-dedicated node.

**Figure 11.** In-and-out connection scheme in case of double POD with EG-dedicated node.

As long as the generation node is well defined, all the requirements for the connection and operation of PGMs in the grid can be applied. Assuming the two POD to be fed by stations of the Transmission Network that are far enough from each other (in an electrical sense), in this configuration, the generator could operate more safely. Moreover, it could operate in line with the prescriptions for its specific category, bearing in mind the intrinsic limitations of the plant evaluated in Section 3.

Of course, from the demand point of view, the influence of the pulsed loads in this case is completely reflected to the transmission system.

#### *4.3. Double POD with PP EPS-Dedicated Node*

Another possible solution is the implementation of a double POD with one node dedicated to the PP EPS. Regarding the SS EPS plus EG EPS node, Figure 12 shows that, both in case of direct and indirect coupling, the maximum power is attained in generation mode, so the requirements to apply for the connection are those referred to the generation nodes.

**Figure 12.** Power profiles at the SS EPS + EG EPS HV node in case of double POD configuration.

In particular, since the maximum power injected is higher than 350 MW, the connection shall be implemented at the highest available voltage level, with the in-and-out double bus-bar scheme (Table 2). In case of indirect coupling, the node appears to grid as a simple generation node, since the power needed for the SS EPS is entirely supplied by the EG EPS during all the operational phases. Therefore, of course the net output at the HV node level is lower than the previous configuration (EG-dedicated node), but the requirements for the connection and operation of PGMs in the grid are still applicable. In case of direct coupling instead, the node is both a generation and demand node since the EG EPS can only supply the whole power required from the SS EPS during the burn time.

Concerning the PP EPS-dedicated POD, the preliminary power profile at the HV node is presented in Figure 13.

**Figure 13.** Power profile at PP EPS-dedicated HV node in case of double POD configuration.

The maximum power absorbed at the node level is higher than 100 MW, so the connection shall be implemented at the highest voltage level available with the in-and-out single bus-bar scheme with bypass.

Figure 14 reports the in-and-out connection schemes in case of double POD configuration with PP EPS-dedicated node.

**Figure 14.** In-and-out connection scheme in case of double POD with PP EPS-dedicated node.

This configuration allows preserving the advantages of the previous case, decoupling both the SS EPS and the EG EPS from the supply of the PP EPS. Since the SS EPS requires practically constant input power, the generator can operate safely, supplying both the internal SS power system and the grid without further stresses and, in case of indirect coupling, as a conventional PGM connected to the grid. At the same time, the power peaks detected by the grid at the PP EPS POD, are of course lower than those related to the previous case where the SS EPS contribution makes them reach higher values.

#### **5. Features of the Point of Delivery (POD)**

This section reports further evaluations on the Point of Delivery (POD) that can be performed starting from the preliminary power profiles elaborated through DIgSILENT PowerFactory [13]. Indeed, talking about the active power profiles, some data can be extrapolated in order to have an idea of the main electrical features of the HV node (or nodes) to which DEMO should be connected. Concerning the reactive power instead, some general considerations are reported.

#### *5.1. Power-Frequency Control*

The requirements in terms of frequency control and admissible frequency variations are defined in the ENTO-E document "P1—Policy 1: Load-Frequency Control and Performance" [14]. The power-frequency characteristic of a power system is linked to its capability of limiting the frequency variation during an event of unbalance between generation and demand. The power-frequency variation is expressed in MW/Hz, so it physically represents the value of power variation in MW that causes a frequency variation of one Hz. Policy 1 [14] defines minimum and average power-frequency control characteristics in Continental Europe, which are respectively 15 GW/Hz and 19.5 GW/Hz. Other important values of the Policy 1 [14] that we have to consider for the purpose of this analysis are the "minimum instantaneous frequency after a loss of generation" and the "maximum instantaneous frequency after a loss of load", which are respectively 49.2 Hz and 50.8 Hz. This means that the maximum frequency deviation accepted in Continental Europe's synchronous area is ±800 mHz from the nominal value. However, it has to he stressed out that these values refer to the loss of a load or of a generation node; they do not refer to the ordinary operation of the grid. Indeed, to have a more reliable limit we should consider the maximum permissible quasi-steady-state frequency deviation that is set to ±200 mHz. This frequency deviation also represents the limit for which all the available

primary control reserves are expected to be fully activated. In fact, this value represents the maximum value that can be managed only relying on the primary frequency control.

Regarding our specific case, analyzing the active power profiles, it is possible to identify several power steps, mainly due to the functioning of the magnets system and of the AH devices. In particular, the most severe event in this sense is a 426 MW active power step that occurs during the Plasma Ramp-up phase. The event, reported in Figure 15, is caused by the PP EPS operation.

**Figure 15.** Active Power step required by the PP EPS during the Plasma Ramp-up phase.

Since this power step is caused by the PP EPS, it will be present in each configuration we considered in the previous section for the connection of DEMO to the electrical grid. Therefore, independently from the solution adopted, we have to evaluate the effect of this event on the grid. Considering the average value of power-frequency control characteristic for Continental Europe, we can calculate the frequency deviation caused by the step in Figure 15, which turns out to be around +21.85 mHz, significantly lower than the 200 mHz limit. Indeed, the 200 mHz limit is reached in case of what is called Reference Incident inside the ENTSO-E documents, i.e., an unbalance between generation and demand of 3 GW. Nonetheless, this event refers to an unbalance condition distributed inside the European transmission grid and not to an unbalance in a single node. In this sense, the effect of a 426 MW active power on the grid step must not be underestimated just because it theoretically does not generate a significant frequency deviation from the nominal value. Moreover, it must be stressed out that this step is not occasional but is cyclically repeated during DEMO operation, so this is another critical aspect.

#### *5.2. Voltage Drop and Short-Circuit Power*

Considering the reactive power, the connection node should be characterized through the short-circuit power. The link between the reactive power and the short-circuit power of the node is the voltage drop. It is well known that, for a fixed value of reactive power required by the facility, a higher short-circuit power of the connection node involves a lower voltage drop at node level. A first approximation of the voltage drop, expressed in per unit, ΔV [p.u.] caused by a variation of reactive power ΔQ on a node with a short-circuit power Ssc is provided by the well-known formula ΔV [p.u.] = ΔQ/Ssc.

Since for now there is not any reliable data regarding the actual reactive power required for the operation of DEMO (mainly for what concerns the supply of the pulsed loads), we cannot really define a specific value for the short-circuit power needed at the connection node. Nevertheless, we can adopt an opposite approach starting from a reasonable value for the short-circuit power and from the requirements in terms of voltage drop defined by ENTSO-E. Bypassing the requirements in terms of power factor and so assuming DEMO to be a non-conventional demand facility, we can preliminarily estimate a reasonable maximum value of reactive power that the facility can demand from the grid.

Considering the studies carried out on ITER [15] and the size of DEMO with respect to its predecessor, the new facility is likely to be connected to a node with a short-circuit power rating in the order of at least 30–40 GVA. This value seems reasonable considering the configuration of the European electrical grid and also considering that nodes with this rating can be also found in Italy [16], where the shape of the country does not facilitate the creation of highly meshed grids (which means higher reliability and also higher short-circuit power).

Bearing in mind this range of values for the short-circuit power, we must define a range of values for the voltage drop that do make sense for the specific case. Considering that most of the reactive power will be required by the PP EPS (due to the power conversion systems), an estimation based on a maximum voltage drop at the connection node, around ± 3 ÷ 5%, should be conservative and reliable. Indeed, the 3% value for the voltage drop is the one assumed for the Pulsed Power Electrical Network (PPEN) in ITER, which corresponds to the PP EPS in our case, as mentioned in the System Requirements Documents of ITER (available in the ITER private database). Due to the considerably greater size of DEMO, it makes sense to assume a wider range for the voltage drop.

Defining the worst case in terms of short-circuit power and voltage drop, i.e., 30 MVA and 3%, respectively, the resulting maximum reactive power demand from DEMO facility is around 900 MVAr.

Unlike the considerations reported for the active power related aspects, in this case the POD configuration affects the evaluations. As already mentioned, the main concerns in terms of reactive power absorption come from the power electronic devices needed to supply the magnets system and the AH and so the pulsed loads. Therefore, if the PP EPS is isolated from the other substations, which means we are adopting the third configuration (see Section 4.3), we can consider only its node as an unconventional demand node, and we can cut the contribution of the SS EPS to the total reactive power required from that specific node. If the first configuration is considered instead (see Section 4.1), since the EG EPS could not be operated as a conventional PGM in any case, the possibility of using it for the internal reactive power compensation is not to be excluded, still bearing in mind the drawbacks of this solution.

#### **6. Conclusions**

Considering the state of the art of the DEMO project, the present work has been carried out with the purpose of providing some cause for reflection on the choices that have been made and that will be made; it is not meant to define final design solutions. In fact, it is important to stress out that at this stage of the design, the data is continuously questioned, so the analyses carried out for the realization of this article aim at providing guidelines, procedures and addresses.

The data available from the studies that are being performed in research centers all over Europe allowed a first estimation of the input and output power profiles of the facility. Through the analysis of these profiles, it has been possible to evaluate the limitations of the facility, both in case of direct and indirect coupling between the PHTS and the PCS, from the generator operation point of view. In this sense, the conclusion is that the direct coupling configuration, despite being the simplest from the constructive point of view, would lead to the impossibility of considering DEMO as a conventional power plant, unless significant adjustments are foreseen. In any case, the eventuality that DEMO could be considered as a non-conventional generation and demand facility is not to be excluded. In this sense, no matter which configuration will be selected, the European grid could allow the operation of the facility. However, this would not be in line with the purpose of the whole roadmap, since it would lead to a single facility that is allowed to operate freely in the grid. If designed on these terms, DEMO would not be a model for real fusion power plant prototypes; it would be a model for a new generation of non-dispatchable generation plants.

Regarding the connection configuration and characteristics, this article presented three possible solutions. In particular, the solution that allows for the implementation of an EG-dedicated point of delivery, even if it is theoretically not in line with the idea at the base of DEMO, which is the self-sustainment of its loads, is the one that shows more advantages. From the grid connection point of view, the generator could be operated in line with the requirements for its category, being the node a pure generation node. From the facility point of view instead, the turbine-generator group would be decoupled from the power spikes absorbed and injected by the PP EPS, guaranteeing a longer life of the mechanical components.

**Author Contributions:** Conceptualization, M.C.F. and M.P.C.; methodology, M.C.F. and M.P.C.; software, M.P.C.; validation, M.C.F., A.L. and S.C.; formal analysis, M.C.F. and M.P.C.; investigation M.P.C.; resources, A.L. and S.C. data curation, M.P.C. and A.L.; writing—original draft preparation, M.P.C.; writing—review and editing, M.P.C. and M.C.F.; visualization, M.P.C.; supervision, M.C.F., A.L. and S.C.; project administration, M.C.F., A.L. and S.C.; funding acquisition, M.C.F., A.L. and S.C. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom Research and Training Programme 2014–2018 and 2019–2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **Electrical Loads and Power Systems for the DEMO Nuclear Fusion Project**

**Simone Minucci 1, Stefano Panella 2, Sergio Ciattaglia 3, Maria Carmen Falvo <sup>2</sup> and Alessandro Lampasi 4,\***


Received: 16 March 2020; Accepted: 22 April 2020; Published: 4 May 2020

**Abstract:** EU-DEMO is a European project, having the ambitious goal to be the first demonstrative power plant based on nuclear fusion. The electrical power that is expected to be produced is in the order of 700–800 MW, to be delivered via a connection to the European High Voltage electrical grid. The initiation and control of fusion processes, besides the problems related to the nuclear physics, need very complex electrical systems. Moreover, also the conversion of the output power is not trivial, especially because of the inherent discontinuity in the EU-DEMO operations. The present article concerns preliminary studies for the feasibility and realization of the nuclear fusion power plant EU-DEMO, with a special focus on the power electrical systems. In particular, the first stage of the study deals with the survey and analysis of the electrical loads, starting from the steady-state loads. Their impact is so relevant that could jeopardy the efficiency and the convenience of the plant itself. Afterwards, the loads are inserted into a preliminary internal distribution grid, sizing the main electrical components to carry out the power flow analysis, which is based on simulation models implemented in the DIgSILENT PowerFactory software.

**Keywords:** balance of plant; DEMO; electric loads; nuclear fusion; plasma; power flow; power supply; power systems

#### **1. Introduction**

EU-DEMO (the DEMOnstration fusion power reactor proposed by the European Union), or simply DEMO, is a unique European project, as it will be the first demonstrative nuclear fusion power plant able to produce and distribute electrical power throughout Europe, thanks to a connection with the European High Voltage (HV) electrical grid (typically at 400 kV) [1–3].

To accomplish this challenging purpose, the European Union set up the EUROfusion Consortium, whose main goals and tasks are summarized in the "European Research Roadmap to the Realisation of Fusion Energy" [4]. The schedule and the milestones of the Roadmap are sketched in Figure 1.

Even though other alternative approaches are being investigated in EUROfusion [4] and in another research facility [5], the EUROfusion Roadmap is based on two tokamak projects: DEMO and ITER [6]. The latter is currently under construction in Cadarache (France) with a worldwide contribution and aims at:

• Producing 500 MW of fusion power for pulses of at least 400 s.


**Figure 1.** Overview of the European Roadmap to the realization of Fusion Energy [4]. The reported dates are still indicative.

These goals are important and partially common with the DEMO project, so that it can use ITER-like solutions for possible future issues. However, DEMO will be bigger than ITER in terms of size and of required services, also because it will be connected to the grid to deliver the produced electrical energy, unlike ITER. Also the time required for the realization of these two projects is different: while it is foreseen to start ITER first experiments by 2025 and to operate with deuterium and tritium by 2035, DEMO is expected to be in operation around by 2050.

In the past, fusion devices were generally not regarded as nuclear facilities and did not need a nuclear license. However, ITER and DEMO are much more critical in terms of tritium inventory, neutron flux, pulse duration, stored magnetic energy, cooling system enthalpy and amount of helium at 4 K. ITER demonstrated its safety and obtained the nuclear license to start the construction. Nevertheless, specific nuclear regulations are likely to be introduced for next-generation devices, also depending on the host country. As DEMO is expected to have more neutronic flux and more inventory of tritium than ITER, a license from authorities will be necessary before starting the construction of safety-classified systems.

As a nuclear facility, a specific design is necessary for the DEMO Balance of Plant (BoP), that is the nuclear engineering term referred to all the supporting and auxiliary systems needed for energy conversion and delivering, excluding all the nuclear components. Therefore, EUROfusion is promoting a multidisciplinary research and engineering activity that is approaching the design of the DEMO BoP [7,8], even moving from the experience of the other experimental tokamaks and nuclear-fission power plants. One of the most critical part of the BoP is the electrical power system, also because this system is not trivial both in terms of size and of complexity.

The present paper introduces the preliminary studies for the feasibility and realization of the electrical system of the nuclear fusion power plant DEMO. In particular, the results of the following activities are described:


The loads characterization and classification is the starting point for first electrical designs and is expected to provide more realistic data than those in previous analyses based on theoretical considerations [11].

This paper is organized in seven sections. Section 2 introduces the relevant DEMO figures and explains possible configurations and operation phases. Section 3 presents the preliminary layout of the DEMO site as used for the electrical analysis. Section 4 introduces the main options of the BoP and the basic principles for the design of the electrical systems. Section 5 is focused on the survey of the main DEMO subsystems and electrical loads and on the results obtained by the load analysis and characterization. Section 6 presents the results on the preliminary design and sizing of a part of the internal distribution grid. Section 7 reports the paper's conclusions.

#### **2. DEMO Characteristics and Operation Phases**

The tokamak operations are based on the heating of a plasma up to temperatures at which it is self-sustained by the fusion processes induced by the ion thermal motion. In DEMO, the plasma heats up the surrounding structure, the tokamak Breeding Blanket (BB), and the fluid used to cool down the BB can drive a Turbine Generator (TG) through proper heat exchangers. The energy produced by such process is expected to be higher than the energy employed to initiate the fusion reactions. In order to reach high temperatures (about 150 M◦C), DEMO could use three different kinds of additional Heating and Current Drive (H&CD) systems [10,12]:


The heat produced by the plasma fusion must be transferred to Power Conversion System (PCS) able to transform it into electrical energy with maximum possible efficiency and reliability: this is the main scope of the BoP. In the present status of the DEMO project, two alternative solutions are considered as basic fluid to cool down the BB: water and helium [8,13,14]. Consequently, two options are considered for the Primary Heat Transfer System (PHTS):


The DEMO operations are strictly related to the physics of the plasma. This also implies a difficulty to achieve very long or steady-state operations that would be preferable for the energy budget. A lot of

fusion research is devoted to the possibility of steady-state operations, but presently without relevant practical results. On the other hand, pulsed operations simplify tokamak physics and may be more flexible in an energy market ruled by renewable sources. Presently, the DEMO operations are supposed to be pulsed in basic option but could be steady-state in future advanced ones. Therefore, unlike nuclear-fission power plants, the DEMO BoP and electrical systems must be designed for pulsed operations for both the two cooling system options in the PHTS-PCS.

In fact, even though a relevant output power is produced only during the plasma flat-top phase, other operation phases are necessary to achieve the correct execution of the pulse. In particular, the DEMO operations consist of seven phases:


This pulsed behavior may introduce specific problems for the BoP. First, discontinuous operations would be damaging to the turbine, then the variable flow of the expected huge powers may let some instabilities arise into the external grid that could even refuse or limit the connection. This problem is even more critical because of the relevant reactive components in power. The durations of the main plasma phases according to last DEMO design are summarized in Table 1.


**Table 1.** Summary of the DEMO operation phases with typical durations.

In order to reduce the output power fluctuations, an intermediate buffer system could be inserted between the PHTS and the PCS. Therefore, two different approaches are under evaluation about the PHTS-PCS coupling [13,14] for both the two cooling options:


The thermal efficiency in the direct cycle option is higher but the TG's life cycle is compromised because it is turned on only during the burn flat-top phase. The thermal energy is extracted from the BB by water or helium, then it is transferred to steam supplying the steam turbine.

In the indirect cycle, the PHTS is coupled with the PCS through the IHTS based on a molten-salt ESS. Its task is to store thermal power during the burn flat-top phase (removed by the molten salt) and its delivery to the PCS. In this way, the TG can operate almost in steady-state at 80% of the PHTS-rated power without interruption nor fluctuations during the plasma phases when fusion heat is not available for energy conversion. Therefore, it can produce almost constant electrical power and rotate at a rather constant speed, thus avoiding thermo-mechanical cycling issues.

Considering the two coupling configurations and the two coolant options, the DEMO BoP could be based on one out of the following four different possible configurations, as summarized in Figure 2:


It is worth noticing the presence of two steam generators in Figure 2b. Unlike the HCPB configuration in Figure 2d, the IHTS in the WCLL BoP takes and stores not all the power coming from the BB PHTS but only the fraction coming from the first wall that is delivered to the PCS during the dwell time using a suitable steam generator [13].

The selection of the optimal configuration is expected to be completed in the next years basing also on the outcomes of the research on the DEMO electrical loads and the design of its distribution network.

(**d**) HCPB indirect cycle

**Figure 2.** The four possible PHTS-PCS configurations that are under evaluation for the DEMO design: (**a**) WCLL direct cycle, (**b**) WCLL indirect cycle, (**c**) HCPB direct cycle, (**d**) HCPB indirect cycle.

#### **3. Preliminary DEMO Layout**

Figure 3 summarizes the preliminary layout that is presently expected for the DEMO site [7,16], mostly based on ITER's one. The actual location of the DEMO site is not yet identified and will be defined also following the outcomes of DEMO electrical analysis and requirements. Nevertheless, it is important in order to assess the electrical distribution layout, the electrical loads and cables characteristics.

**Figure 3.** Preliminary layout of the EU-DEMO site.

#### **4. Basic Principles for the Design of the DEMO Electrical Power System**

DEMO power systems can be divided in three electrical distribution groups:


Figure 4 shows a preliminary sketch of a possible configuration of the DEMO electrical power system with its three sub-distribution groups.

The scheme in Figure 4 should be integrated by emphasizing the buses for the safety loads and by inserting the systems for the reactive power compensation and harmonic filtering. Such systems are relevant in ITER (occupying an area approximately corresponding to the DEMO Area 35 in Figure 3), but their ratings and placements can be defined in DEMO only after an adequate survey of the electrical loads.

Because of the deeply different nature of the electrical loads connected at SSEN and PPEN distributions, the former can be exhaustively characterized into an electrical load list, as described in the rest of the paper, but the latter need further specifications in terms of time evolution of the power profiles. Such difference heavily affects also the power profiles at the Point of Delivery, requiring further design and optimization stages to meet the requirements imposed by the grid regulator.

**Figure 4.** Preliminary sketch of a possible DEMO electrical power system configuration, emphasizing the division of the electrical distribution into three groups: SSEN, PPEN and TG.

The amount of electrical power delivered by DEMO to the external grid depends on the selected option for the PHTS-PCS and on its control strategy. Figure 5 shows the input and output of the DEMO electric power depending of the PHTS coupling. The figure is essentially qualitative as PPEN contribution is still under analysis, but some data available for a non-optimized coil system show that it could require input powers at breakdown even higher than 1 GW with relevant reactive power components [15]. More in general, the input power requested could exceed the produced power in some time intervals.

**Figure 5.** Qualitative (not in scale) curves of input (red line) and output powers. The thermal power produced by the plasma and fusion processes (purple line) can be transformed in an electrical power whose characteristics are different for the direct (green line) and indirect coupling (blue line).

With the direct PHTS-PCS coupling, the power for the DEMO electrical loads should be supplied only by the external grid during the dwell time between the pulses and during the ramp-up and ramp-down phases too.

In order to avoid excessive mechanical and thermal stresses on the turbine, it is supplied with a proper mass flow rate of steam which has to guarantee the turbine operation at 10% of its nominal capacity during the dwell time. In order to cover the steam production during the dwell time, the most suitable and efficient solution consists in the introduction of a small thermal ESS loop, managed in one of the following three main configurations:


The most efficient solution seems to be the third one due to temperature concerns [17]. Indeed, using the steam in the PHTS to heat the molten-salt hot tank does not allow reaching a temperature high enough to guarantee proper conditions for the steam turbine during the dwell time. This could result in water damage to the steam turbine.

Besides the operation during the dwell time, thermodynamic analysis show that the maximum power output of the turbine is evaluated to be around 790 MW [18]. Considering a conservative efficiency of 0.95 for the turbine-synchronous generator coupling, the nominal active power to consider for the generator is 750 MW. In literature, a common minimum power factor to operate synchronous generators results to be 0.8. This value is also adopted in some grid legislations [17] to identify the value of the reactive power that a generator should be able to supply to the grid when required, and so to build a reference capability curve. Considering this value for the power factor, the rating of the synchronous generator in terms of apparent power is around 940 MVA.

This has to be considered an approximation since the variations in the prime mover power would actually cause transient phenomena at the synchronous machine level. From the reactive power point of view instead, the nominal value that the generator could supply to the grid is around 565 MVAR, coming from the minimum power factor assumed. However, no assumptions can be made on the reactive power profile since, in general, it is dependent on the conditions of the external grid and so it is managed by the transmission system operator.

In order to find a coherent value for the output power and to assume an efficiency for the new thermodynamic cycle, it is necessary to know the thermal energy balance. Of course, this efficiency might be different from the direct coupling case because of the presence of the IHTS. Using the IHTS, the turbine operates constantly at its nominal power, increasing the overall efficiency during the whole operation (no transients, no off-design conditions). On the other hand, the presence of a more complex heat exchange system may lead to less thermal power transferred to the PCS. For these reasons, a first approximation about the power output in case of indirect coupling has been evaluated as the total energy output during one pulse, divided by the total duration of the pulse. In this way, the nominal power of the turbine has been calculated to be around 85% of the previous case. This leads to a rated active power of around 640 MW. Taking as a reference for the power factor the value adopted in the previous case, for the indirect coupling configuration, the rated apparent power of the synchronous generator ranges around 800 MVA.

#### **5. Summary of Electrical Loads Connected to the SSEN**

The DEMO subsystems are organized according to a Plant Breakdown Structure (PBS), used also for the classifications of the electrical loads. Table 2 provides for each PBS a short description, the expected power and the electrical distribution system where the PBS and its loads is connected. Since some PBSs contain both loads connected to the SSEN and to the PPEN distribution networks, they are reported twice.


**Table 2.** EU-DEMO Project Plant Breakdown Structure (PBS) with expected powers and connection to the distribution grid.

<sup>1</sup> The PBSs referred to the HCPB and WCLL options are alternative: once the final configuration will be selected, only one of the two reported powers will be requested in DEMO. <sup>2</sup> The powers reported for the H&CD systems (PBSs 30, 31 and 32) were estimated for the reference solution with related efficiencies (see Section 5.3). <sup>3</sup> Since the powers absorbed by these PBSs are not constant, their peak powers are reported. While the PPEN loads (PBSs 30, 31, 32 and 82) may be very variable, the load of PBS 50 is reduced to 20% in some phases as described in Section 5.5. <sup>4</sup> Even if RM System is connected to the SSEN, it mainly operates during specific maintenance phases and not during plasma phases in Table 1.

Passive systems are those not absorbing any significant electrical power. However, it is worth mentioning that systems which are labelled as passive may require a small amount of power for the instrumentation and control systems. In such cases, this power is accounted in PBS 40 or 85.

Some of listed PBSs of the project are still at very preliminary stage or under design. To provide a first estimation of the electrical power absorbed by these subsystem, and to carry out a comprehensive (although preliminary) power flow analysis of the DEMO plant, a first guess based on reasonable extrapolation from the ITER loads and power profiles was used.

For each of the PBSs listed in Table 2 the information necessary to run power flow analyses were investigated, as the load type (if belonging to SSEN or PPEN distribution), the voltage level and the power absorption in terms of active power and power factor. Whenever no information coming from the design layout regarding the power factor were available, a local reactive power compensation system is supposed for each subsystem, regulating it at a value equal to 0.95.

Note that the total input power requested by DEMO cannot be estimated only by summing the entries in Table 2 because: (i) HCPB and WCLL cooling options are alternative and (ii) the power absorbed by loads belonging to PBSs 30, 31, 32, 50 and 82 are not constant for all the phases, being them pulsed. In these cases, the only peak values of the power are reported. In particular, the PPEN loads require high powers only to initiate (phases 1 and 2 in Table 1), ramp (phases 3 and 6) and heat (phase 4) the plasma, while their contribution is lower than the output power for most of the operational time (phase 5). The details for the PPEN power profiles will be developed in next years.

The following subsections provide a brief description of the electrical loads for each "non-passive" DEMO PBS supplied by the SSEN distribution.

#### *5.1. PBS 22: Tritium, Fueling and Vacuum*

The DEMO PBS 22 collects all the components belonging to the subsystems devoted to the fuel cycle (e.g., isotope separation columns, water detritiation, molecular sieve beds), to the plasma fueling (e.g., pellet injection systems) and to the vacuum generation inside the plasma chamber (e.g., motor pumps and compressors). For all these subsystems, preliminary assumptions and designs are under evaluation based on the lesson learnt from ITER layout and projects. A first guess of the total electrical power absorbed by the three subsystems is provided as 12.2 MW, from the information collected from the design progresses and from extrapolations from ITER, taking into account also the differences between the two projects and how they reflect on the operation of each subsystem.

#### *5.2. PBS 25*/*27: Tritium Extraction and Removal for HCPB*/*WCLL*

Tritium Extraction and Removal is a subsystem devoted to the unburned and unexhausted tritium extraction from the VV and its transportation to the fuel cycle system (PBS 22-1). Afterwards, it is treated, separated from its isotopes and re-injected inside the chamber to participate to the occurring fusion reactions. This system is at embryonic state and no information regarding its structure and its power absorption have been evaluated at present, both for the two cases of cooling system based on helium or water coolant. As a first guess of the power absorption, an extrapolation from ITER layout was carried out, obtaining an estimation of 3 MW.

#### *5.3. PBS 30, 31 and 32: Heating and Current Drive (H&CD) Systems*

The H&CD systems foreseen in DEMO are ECRH, ICRH and NBI. The optimal heating source mix, as well as the power to be delivered to the plasma, is still a key issue under investigation. The definition of the final parameters is scheduled by the end of the conceptual phase, foreseen by the end of 2024. Up to now, the requests in terms of plasma heating and stabilization require to deliver 150 MW of heating power to the plasma (that needs a much higher electrical power to be supplied). Three possible options are under investigation to achieve the total amount:


It is important to stress that the functions of the H&CD systems need to be supplied by two different kinds of power supplies:


In ITER the auxiliary powers of the H&CD systems are included in the respective PBSs and are accounted to be all equal to about 3 MW. As the DEMO power supplies are expected to manage higher powers, an auxiliary power of 6 MW is estimated from the SSEN for each of the three H&CD PBSs.

The issue of the time evolution and efficiencies [10,12] of the H&CD Main Power Supplies is more complex and will be addressed in a future paper by the authors. The average efficiencies for the expected technologies are foreseen to be about 40% of the power delivered to the plasma [12], resulting in a power of 125 MW for the main power of each H&CD system.

#### *5.4. PBS 40: Plasma Diagnostics and Control Systems*

Since the Plasma Diagnostic & Control System is still to be defined, the electrical power absorbed has been estimated by the ITER data. However, since the DEMO facility is expected to be developed under a more mature knowledge of fusion physics and technology with respect to ITER, its diagnostic system is supposed to be much simpler, leading to an estimation of 6.1 MW, supposing a global power factor equal to 0.9. The electrical power includes the local air conditioning of the cubicles serving the diagnostics.

#### *5.5. PBS 50: BB Primary Heat Transfer System (HCPB)*

The Primary Heat Transfer System based on the HCPB collects all the components and systems devoted to the fusion heat recovery and delivery to the PCS. The main heat extraction zone is the BB, supplying about 85% of the total fusion heat. Other fusion heat sources such as the VV, limiter and divertor are addressed in PBS 49 and PBS 58. As mentioned in Section 2, two possible layouts are under investigation to couple PHTS to the PCS, in order to mitigate the intrinsically pulsed behavior of the plasma for TG, that are the indirect (Figure 2d) and direct coupling (Figure 2c).

In the first configuration, the electrical loads are related both to the PHTS and to the IHTS. Its aim is to recover heat to be supplied to the turbine during dwell time and all phases when fusion heat is not available for energy conversion, in order to keep it constant in its operation.

In the direct cycle configuration, instead, the electrical loads are only those belonging to the PHTS system and those belonging to the small ESS.

Considering the segmentation option with 18 sectors (each of them including three outer blanket and two inner blanket segments), the HCPB BB is divided into nine independent circuits (for safety reasons, in order to limit common-mode failures), serving two sectors each. The sectors are fed by nine loops (each with two compressors), six designed for the Outboard Blanket (OB) sectors and three designed for the Inboard Blanket (IB) sectors. The main electrical load for each loop is the electrical motor connected to the compressors, for which a first power estimation is provided based on the thermo-hydraulic design of the loops themselves (9.2 MW of mechanical power per compressor with efficiency equal to 0.95).

Because of the very high electrical power absorbed by the motor compressors, resulting in a total power in the order of hundreds of megawatts, it was decided to regulate them in order to reduce the power consumption during the scenario phases when no thermal power comes from the fusion reactions. Unfortunately, the power transitions from regulated to full power and vice versa cannot be performed at any time rate because the current technology for helium compressors allows a ramp-up and ramp-down at a maximum rate of about 10–20%/minute. Also with controlled ramps, the frequent power variations induce stresses on the compressors that are justified by the relevant energy saving that can be achieved (more than 100 MW for more than 500 s). Some possible modulations of the compressor power are sketched in Figure 6. The selected modulation rate will be a trade-off between energy saving and components stress. Presently, the assumed tradeoff is 20%/min but faster rates will be investigated also considering technology progresses.

**Figure 6.** Possible qualitative time evolutions of the power modulation of each of the 18 HCPB OB and IB BB compressors in order to reduce the total power demand. The assumed tradeoff is 20%/minute, a slower rate makes no sense, a faster rate could be detrimental for the components.

In the IHTS there is an intermediate heat exchanger based on "tubes and shell" technology: helium flows into the tubes and molten salt crosses the shell side [13]. In order to mitigate the dynamic behavior of the thermal fusion power between burn flat-top and dwell time on PCS components (particularly on TG and then on the electrical grid), the IHTS is equipped with an ESS. Its task is to store thermal power during burn flat-top phase (removed by the molten salt) and its delivery to the PCS in order to let the TG work almost in steady-state at 80% of the PHTS-rated power, thus avoiding thermo-mechanical cycling issues. Its main electrical loads are the electrical motors moving the circulators present in the loops, whose electrical power was estimated from the thermo-hydraulic design of the IHTS system in terms of mass flow rate and pressure drops. Besides, many other electrical loads regarding all auxiliary systems are still to be evaluated both belonging to the PHTS and the IHTS system. Therefore, they are not included in the present estimations since currently under design.

#### *5.6. HCPB PBS 52: Primary Heat Transfer System (WCLL)*

The PHTS based on the WCLL collects all the components and systems devoted to the fusion heat recovery and delivery to the PCS. The main difference with respect to the HCPB architecture is that the heat exchanger is not between PHTS and PCS in the indirect coupling configuration but it receives and stores only the portion of thermal power coming from the first wall of the vessel. This solution aims to reduce the size of the IHTS and its ESS but needs the presence of a second steam generators between the PHTS and the IHTS. However, the same considerations already mentioned for the HCPB hold for the WCLL cooling system about the coupling between the PHTS and PCS as well as the electrical loads to be included in the power flow analysis.

The PHTS consists of two different cooling systems, removing heat from two different parts of the blanket, namely the breeding zone and the tokamak first wall (addressed in PBS 59) and the VV (addressed in PBS 49). Both PHTSs consist of two loops, in order to limit the size of piping and components. The breeding-zone PHTS loops include a steam generator, two pumps, one pressurizer and one hot/cold leg. The first-wall PHTS loops include one heat exchanger, one pump, one pressurizer and one hot/cold leg. Cold/hot ring headers supply/return water to the breeding-zone/first-wall loops. Summing all up, six main coolant pumps are foreseen in the present PHTS WCLL design: four main coolant pumps, two per loop, are present in the cold legs of the breeding-zone PHTS and two main coolant pumps in the first-wall PHTS.

The electrical loads are the electrical motors moving the pumps as well as the pressurizers. Conversely from the motor pumps, which absorb constant electrical power during the whole scenario,

the pressurizers are "time-random" electrical loads. They are resistors benches which heat the water up in order to increase the water pressure in pipes whenever the control system sense a pressure reduction. Therefore, they may or may not be switched on during the scenario and not at the same electrical power. In practice, the total power of the pressurizer is 2 MW in standard conditions but when the pressure drops below the minimum threshold the pressurizer is switched on, demanding 10 MW. The very conservative worst-case assumed in these analyses consists in assuming the pressurizers working continuously.

As shown in Figure 6, in the HCPB option, the helium pumps are modulated to operate at reduced power around the dwell time. This could be theoretically useful also for the PHTS-WCLL, but the limited energy saving that could be achieved does not justify the cycling stress produced on the pumps by the frequent power variations.

The WCLL IHTS is conceptually similar to that of the HCPB BB in terms of thermo-hydraulic architecture. First estimations regarding the electrical power absorbed by the electrical motors moving the circulators present on the loops are based on the estimations of mass flow rate and pressure drop in the pipes.

#### *5.7. PBS 60: Remote Maintenance (RM) System*

In new-generation tokamaks, Remote Maintenance (RM) is becoming more and more demanding and important with respect to existing facilities in terms of electrical power demand and nuclear safety issues. In fact, in ITER and DEMO the Occupational Radiation Exposure is expected to be twice or three times that of modern nuclear fission plants.

The DEMO RM is mainly divided into two groups:


The concepts of the safe transportation of extracted parts of the VV are similar to those adopted in ITER. In fact, once extracted from the vessel, both the divertor and the blanket sectors are moved by transport casks through gallery corridors.

Besides, many crane-based systems and lifting systems need to be foreseen in designing the handling of the extracted BB (through the upper port) and divertor (through the lower port) sectors, to move them to the Active Maintenance Facility. Some megawatts of electrical power are expected to be absorbed by this system based on the first estimations on the weight of the parts to be moved.

The total DEMO maintenance and safety needs can be sustained by a RM of about 5 MW. Even though the RM electrical loads belong to the SSEN, parts of them are "decoupled by the plasma operation", as for the remote handling inside the VV. This is a key point in the design of the electrical distribution: the coincidence factor of many electrical loads is zero. However, it is worth mentioning that some RM activities such as the processing/detritiation of some components could be carried out both during and off operations in the Active Maintenance Facility (Building 21 in Figure 3).

#### *5.8. PBS 70: Balance of Plant (HCPB)*

Components belonging to PBS 70 are those of the PCS with HCPB-PHTS. This system consists of a classical Hirn Cycle with superheated steam, with steam generator, re-heater, deaerator, condenser, feed water and turbine (connected to the electrical generator for the thermal-electrical energy transformation).

First studies about absorbed electrical power by PCS in case of helium cooled blanket only regard the water pump, based on the requested mass flow rate inside the circuit and the estimated pressure drops. Further investigations will regard ancillary systems serving the PCS, not available at this stage.

It is worth noticing that the net efficiency of the thermal cycle is about 31%, when taking into account the electrical power absorbed by PHTS (in particular, the very high power absorbed by the circulators), IHTS and PCS components. The previous considerations lead to a power estimation of 12 MW for PBS 70.

#### *5.9. PBS 72: Balance of Plant (WCLL)*

As for PBS 70, the components belonging to PBS 72 are those of the PCS in the cooling system layout based on the WCLL concept. The same considerations already reported for PBS 70 were proposed, except for slight differences in terms of the steam generator coupling the PCS and the IHTS in case of indirect coupling. It results in a total power absorbed equal to 12 MW. It is worth noticing that first simulations pointed that the system would be able to operate with about 700 MW of gross electrical power and an efficiency equal about to 34%, considering all electrical power needed to be recirculated for PHTS, IHTS and PCS components needs.

#### *5.10. PBS 81: Cryoplant and Cryodistribution*

A preliminary design of cryoplant and cryodistribution is under development both in terms of layout of the plant and of estimation of the power and energy absorbed by the mechanical cryopumps. Among the numerous systems to be kept at cryogenic temperatures, it is worth mentioning: the superconducting coils (at 4 K), the coil casings and supporting structures and the thermal shields of VV, cryostat, ports and cryodistribution. Besides, differently from ITER cryoplant, some systems are foreseen to be refrigerated by independent systems, such as cryopumps for the VV and the divertor, NBI cryopumps, pellet launching system and ECRH superconducting magnets. However, the project is currently at a very preliminary stage. A first guess of the total electrical power absorbed by the DEMO cryoplant and cryodistribution (101.8 MW) has been obtained by scaling that designed for ITER project taking into account the volumes of the superconducting magnets and of the other structures and components to be kept at cryogenic temperatures.

#### *5.11. PBS 83: Buildings*

The electrical loads included in such estimation regard the lighting systems, the elevator systems and the Heating, Ventilation and Air Conditioning (HVAC) system. Since the design of the DEMO buildings is at a very preliminary stage, a first guess of 54.8 MW is provided by extrapolating the data available from ITER and considering the DEMO preliminary site layout shown in Figure 3.

#### *5.12. PBS 85: Plant Control System*

A first guess of 3.6 MW for the power absorbed by the COntrol, Data Access and Communication (CODAC), Central Interlock System (CIS) and Central Safety System (CSS) systems was obtained by estimating the number of cubicles for each subsystem, from those present in ITER.

#### *5.13. PBS 87: Auxiliaries*

First preliminary information about the power absorbed by auxiliaries is an extrapolation from ITER's electrical demand for the Component Cooling Water System (CCWS), CHiller Water System (CHWS), sulfur hexafluoride (SF6) distribution system and compressed air distribution system. This extrapolation results in 90.9 MW.

#### *5.14. Summary of the Total Electrical Power Absorbed by the DEMO SSEN*

Since the BoP is still under design also in terms of choice between the two helium-based and water-based cooling system, two parallel and independent designs are under evaluation also in terms of the power electrical system. Figure 7 summarizes the SSEN active and reactive rated powers in each plasma phase for both the two cooling system options. It is important to note the power variations in the HCPB option (Figure 7a) due to the ramp modulation of the compressors motor in the PHTS circuits.

(**b**)

Active power (MW) Reactive power (MVAR)

**Figure 7.** Nominal SSEN active and reactive powers in each plasma phase, considering the two different PHTS-PCS options: (**a**) HCPB and (**b**) WCLL.

#### **6. Preliminary Design and Sizing of the DEMO Electrical Distribution**

In compliance with the preliminary scheme in Figure 4 and the lesson learnt from ITER electrical design, four voltage levels are firstly assumed to be used for the DEMO electrical distribution system:


According to this assumption and to the aforementioned loads characterization, the HV/MV and MV/LV transformers and the cables for SSEN are sized, by using simulation models for the power flow analysis [19]. The electrical scheme has been designed and implemented in the DIgSILENT PowerFactory software environment. The criteria used for the simulations are:


The flow diagram in Figure 8 summarizes the algorithm adopted for the design of the electrical distribution system and for the sizing of the electrical components.

**Figure 8.** Simplified flow diagram of the algorithm adopted for the preliminary design of the DEMO SSEN.

Simulations were carried out independently for both the two different cooling system configurations (PHTS-HCPB and PHTS-WCLL) and referring only to SSEN loads in each plasma operational phase. As first output of the power flow analysis, the number and the size of the HV/MV transformers are identified according to available data of installed transformed and reliability issues both during normal and off-normal conditions: six three-phases 2-windings transformers with rating power of 150 MVA and voltage ratio equal to 400/22 kV.

As regards the internal distribution for SSEN, the loads have been dispersed following design criteria based on the voltage levels, loads type and distance between the distribution node and the actual location of the loads in the facility estimated on the site layout. The ordinary and the safety loads are distinguished at the 0.4 kV voltage level and the following MV/LV transformers are selected following the same aforementioned criteria:


The choice to divide the MV/LV transformers in several groups connected by a bus-tie aims at limiting the short-circuit currents, thus allowing the selection of commercial low voltage circuit breakers. Regarding the loads connected at the 6.6 kV busbars, they are supplied through:

• Five three-phases 2-windings transformers with rated power of 40 MVA and voltage ratio equal to 22/6.6 kV in case of HCPB.

• Five three-phases 2-windings transformers with rated power of 20 MVA and voltage ratio equal to 22/6.6 kV in case of WCLL.

In the 6.6 kV section, the ordinary loads are supplied by 10 three-phases 2-windings transformers with rated power of 40 MVA and voltage ratio equal to 22/6.6 kV. Finally, the cryogenic system is supplied by five three-phases 2-windings transformers with rated power of 40 MVA and voltage ratio equal to 22/6.6 kV. The main data of the selected transformers are reported in Tables 3 and 4, where:



**Table 3.** Electrical data of the MV/MV and MV/LV transformers.


Results obtained by the methodology based on the power flow analysis and the aforementioned sizing criteria allowed also to carry out the cables sizing. It is important to stress that this sizing was possible only by adopting a preliminary but consistent layout as shown in Figure 3. The type of the selected cables and their electrical data are summarized in Table 5, reporting the value of the resistance *R*' (for the reference and maximum operating temperatures) and reactance *X*' per unit length. The assumed cable materials and technical characteristics are summarized in Table 6.



**Table 6.** Materials and technical characteristics assumed for the SSEN cables.


From the power flow analysis carried out including the SSEN loads and distribution, the total electrical active and reactive powers requested to the HV point of connection of DEMO to the external electrical grid can be assessed in reference to the different plasma phases. The results are shown in Figure 9 for the HCBP and WCLL cases. This information is the starting point for the sizing and design of the HV switchyard that is object of study and research of the same authors.

**Figure 9.** Total steady state electrical powers requested from the grid at the DEMO HV connection point in the two PHTS-PCS cases (HCPB and WCLL).

#### **7. Conclusions**

This paper presented the preliminary studies for the feasibility and realization of the demonstrative nuclear fusion power plant DEMO. The study moved from the analysis of the expected DEMO operations and of the electrical loads to sketch a preliminary electric scheme for a part of the internal distribution grid, including the sizing of the main electrical components. This preliminary design was supported by simulation models for the power flow analysis implemented in the DIgSILENT PowerFactory environment.

The first results are focused on the electrical loads analysis and characterization showing a significant variability with respect to the plant configuration, especially because of to the cooling choices and the operational phases. This aspect affects the preliminary design of the electric distribution grid for the SSEN in DEMO. The results show a huge amount of power that would be shared with the HV European grid, pointing out the complexity of the sizing and design of the HV switchyard.

The results shall be continuously refined in the next years following the development of the project and the new experiences in the plasma physics and engineering achieved by ITER or satellite experiments. For instance, the design optimization of the toroidal and blanket segmentation led to a number of sectors being reduced from 18 to 16. The following design choices could lead to a piping redistribution with a limited impact on the total pumping power but there is still room for general design optimizations. Moreover, a relevant breakthrough could consist in the possibility to achieve longer burn phases.

Nevertheless, the electrical design must be continuously updated because its impact is so relevant that could jeopardy the efficiency and the convenience of the plant. In fact, some physics and engineering choices as the cooling configuration and the coil current scenarios could be driven also by electrical constraints or optimizations.

In particular, the following aspects are object of study and research in progress by the same authors:


**Author Contributions:** S.M. collected the data from DEMO PBS experts and from other tokamaks. S.M. and S.P. modeled the distribution network and performed the power flow simulations. S.C. followed the conceptualization of the BoP, of the layout and of the safety classifications. S.P. conceptualized and developed the electrical network. A.L. arranged the electrical load list and the operational phase classification. The methodology was agreed by all the authors. S.M., S.C., M.C.F. and A.L. organized and supervised the project for the involved institutions. The manuscript was edited by all the authors. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **Abbreviations and Acronyms**


#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

*Article*
