Case studies of small pumped storage

ABSTRACT

Energy storage through pumped-storage (PSP) hydropower plants is currently the only mature large-scale electricity storage solution with a global installed capacity of over 100 GW. The objective of this study is to evaluate the possibility of using this storage solution on a smaller scale to provide local voltage control and line congestion management to active medium-voltage distribution networks. First, several potential sites were identified in the French-speaking part of Switzerland, ideally with existing reservoirs. Following a selection of the most promising sites, two are selected to pre-size the various components and estimate the investment costs. These two sites are located in the Val de Bagnes in the canton of Valais in Switzerland and offer hydraulic powers of approximately 5 and 10 MW. The approximation of the investment cost is around 2 CHF/W. In comparison with recent large hydropower projects these prices seem quite reasonable, given that the investment cost for large hydropower is on the order of 2–3 CHF/W. In parallel, a free software (“freeware”) program to evaluate the potential, to select the components and to estimate the costs is developed to facilitate future studies.

RESUME

Le stockage d’énergie grâce aux centrales hydroélectriques de pompage-turbinage est actuellement la seule solution mature de stockage d’électricité à grande échelle avec une capacité installée mondiale de plus de 100 GW. L’objectif de cette étude est d’évaluer la possibilité d’utiliser cette solution de stockage à plus petite échelle pour assurer le contrôle local de la tension et la gestion de la congestion des lignes aux réseaux de distribution. Dans un premier temps, plusieurs sites potentiels ont été identifiés en Suisse Romande idéalement avec des réservoirs existants. A la suite d’une sélection des sites les plus prometteurs, deux sont sélectionnés pour en pré-dimensionner les différents composants et estimer les coûts d’investissement. Ces deux sites se trouvent dans le canton du Valais en Suisse et offrent des puissances hydrauliques de l’ordre de 5 et 10 MW. L’approximation du coût d’investissement est de l’ordre de 2 CHF/W. En comparaison avec des projets récents dans la grande hydraulique ces prix semblent tout à fait raisonnables sachant que leur coût d’investissement est de l’ordre de 2–3 CHF/W. En parallèle, un logiciel gratuit (« freeware ») pour évaluer le potentiel, sélectionner les composants et estimer les coûts a été développé pour faciliter les études futures. 

KEYWORDS:

• Investment cost
• renewable energy
• energy storage

MOTS CLES:

• Coût d’investissement
• énergie renouvelable
• stockage d’énergie

1. Introduction

In October 2017 Switzerland ratified the Paris Agreement aiming to cut greenhouse gas emissions to net zero by 2050. In January 2021 the Federal Council released its roadmap to achieve the Paris Agreement by setting 10 key strategic principles that will guide and shape Switzerland’s climate policy actions over the coming years (The Federal Council, 2021). This implies that most of the present sources of energy, such as fuel oil, fuel gas and nuclear power, will need to be replaced by renewable energies such as solar and wind energies. However, in 2018 these two energy types represented, respectively, 2.9% and 0.2% of the annual electricity production. Therefore, a significant effort will be required to find an alternative to non-renewable energy, which accounts for approximately 40% of energy generation in Switzerland (OFEN, 2020). The expected alternative relies on solar energy with a potential photovoltaic capacity of 37.5 GW (Swissolar, 2020).

The usage of wind and solar energies has the well-known problem of intermittency. The amount and the time of production cannot easily be predicted. Furthermore, the peak of production might not coincide with demand, making its management more complicated. Therefore, a storage solution should be used to ensure that excess energy is not lost. One solution is to use a battery, as demonstrated when Tesla installed a 100 MW/129 MWh lithium-ion battery connected to a wind farm in Hornsdale in the state of South Australia (IRENA, 2020). However, such technology is, as of today, still under development and suffers challenges such as limited cycle life, cost, and poor performance in hot and cold climates (Kabir & Demirocak, 2017).

The other storage alternative is the well-advanced pumped-storage technology. Two reservoirs at two different altitudes will act as a battery. The excess of energy will be converted into mechanical energy via a pump and used to transfer the water from the lower reservoir towards the upper one, thus giving the water potential energy. Then, when the energy demand exceeds the supply, the water from the upper reservoir can be released to the lower one. Its potential energy will be converted to kinetic, mechanical and, finally, electrical energy. However, this highly efficient and reliable technology requires water reservoirs. Martínez-Jaramillo et al. (2020) analysed the feasibility of 100% renewable generation in Switzerland. They considered hydro and photovoltaic generation combined with pumped-storage hydro. Their analysis showed that the pumping capacity should be doubled, and the reservoir size increased by up to 100% depending on the installed solar capacity. It is, therefore, necessary to find a new approach to increase the pump and storage capacity at a reasonable cost.

This paper focuses on the development of a small-scale and affordable pumped-storage powerplant. To do so, a survey was conducted to find existing reservoirs, such as the ones used for artificial snow or irrigation, in the Vaud and Wallis cantons in Switzerland (Martignoni et al., 2018). From this research several sites are found to be feasible and two of them are selected. A feasibility study is carried out to evaluate the future components of the powerplant as well as the required investment cost.

2. SIG COGENER project

The SIG (Services Industriels de Genève) COGENER (comité genevois pour l'utilisation du Fonds SIG pour les Nouvelles Energies Renouvelables) project, led by the HES-SO Valais-Wallis (Haute École spécialisée de Suisse occidentale) and Mhylab from July 2015 to June 2017, aimed to assess the relevance of using small-scale pumped-storage powerplants as a possible solution to regulate the production of intermittent energy such as solar and wind energies. This assessment focused first on an analysis of the hydraulic potential of two cantons located in the French-speaking part of Switzerland; second, on the identification of an economical model for this type of service; and, finally, on clarifying the relevance of developing this new technology suitable to equip small- to medium-head sites and thus increase the installed power and storage capacity.

The first phase of the project identified the needs in terms of storage energy, pumping and generating mode durations as well as the power of the installation. These identifications are based on interviews with transmission system operators (TSOs), distribution network operators (DNOs), balance group managers (BGMs) including BGMs for renewable energy, and potential customers interested in this new type of technology. It was found from these interviews that an interest exists in systems for energy storage by small-scale pumped-storage. The main usage of this new storage would be in mitigating the power peak resulting from the start of the industry or from human activity. Therefore, the ideal power would be between 1 and 10 MW, to remain in a small hydro area, with a pumping and generating mode duration from 2 to 6 hours and a powerplant usage from 1 to 2 cycles per day.

Two other usages were also foreseen. The first was an ancillary service with primary control. This service would require a power level between 5 and 10 MW with 4 to 6 cycles per day. The second was also an ancillary service but for the secondary control. A power level between 200 kW and 5 MW would be required with a storage capacity allowing several cycles of 15 minutes per day. However, to make this technology competitive, it is necessary to integrate the future storage system into an existing structure such as the irrigation, drinking water or mechanical snowmaking networks, and thus provide them with additional flexibility.

Following the interviews, the technical, electrical and functional characteristics of the storage system were established. Based on these characteristics, an initial census was carried out. This initial census consisted of listing the existing reservoirs with a sufficient hydraulic potential. These reservoirs considered can be natural or artificial lakes, used for turbine operation, for artificial snow, for irrigation, for drinking water or even for flood prevention (Crettenand, 2012).

The initial census was carried out in the cantons of Vaud and Wallis in Switzerland and resulted in the identification of 186 natural lakes or artificial reservoirs. The number of reservoirs might be higher as not every drinking water network was analysed. From these 186 reservoirs, 19 were identified for the installation of a small-scale pumped-storage powerplant, spread over 27 exploitable reservoirs. Thus, the total technical potential is estimated at more than 75 MW, for a storage capacity of around 430 MWh. The situation of the 186 reservoirs identified is shown in Figure 1. The 27 potential reservoirs for small-scale pumped storage are highlighted in dark blue. Among these 19 potential sites, two attracted the attention of local authorities and were analysed in more detail. These sites are located in Valais in the Bagnes Valley, and are discussed in detail in the next section.

Figure 1. Potential sites to instal a small-scale pumped-storage powerplant (Martignoni et al., 2018). The disc location and size represent the location of a potential reservoir and its volume, respectively. The darker blue colour emphasises the reservoirs that already have a potential for usage.

To facilitate the study of a small pumped-storage power plant, an in-house software program was developed using Python 3.7 and the PySimpleGUI library (version 4.18.2). The results presented in the next section were obtained using this program. The different cost models were developed based on the literature or quotes obtained from suppliers.

3. Case studies

3.1. Site Louvie lake–Fionnay reservoir

3.1.1. Site description

The site is located in Fionnay in the Wallis Canton, where two reservoirs already exist. The upper reservoir, called Louvie, is an artificial hillside catchment reservoir used for irrigation and drinkable water in the Bagnes Valley. It is also used for fishing and is located at an altitude of 2214 m. It was built in the 1960s and has a water capacity of 340,000 m³. As of today, the water exploitation is given to the company called ALTIS. The lower reservoir is a compensation basin used by the “Forces Motrices de Mauvoisin SA” (FMM) to compensate for the difference in maximal turbine discharge between the powerplants located in Fionnay and Riddes. Its water capacity is 170,000 m³ and it is located at an altitude of 1491 m. The locations of the two reservoirs are shown in Figure 2. For the present analysis, it is assumed that 10% of the water capacity of the compensation basin can be used. Based on this assumption, the energy storage E (kWh) can be computed as follows:

(1) $E=2.78\cdot {10}^{-7}\cdot \rho gVH$(1)

Figure 2. Geographical location of the Louvie–Fionnay reservoirs (red surfaces). The preliminary piping path connecting the two reservoirs is shown with a blue line. Source: Swisstopo.

where ρ (kg m⁻³) is the water density, g (m s⁻²) is gravity, V (m³) is the available water volume and H (m) is the head. The capacity of energy storage for this site reaches 33.4 MWh, providing a hydraulic power of approximately 5 MW based on a cycle of approximately six hours in generating mode.

Regarding access, all equipment can easily be brought by road to Fionnay. The water intake at Louvie, the settling basin if necessary, and the pipes would be the most complex to instal as it is located in a mountainous area. Helicopters will certainly be necessary.

3.1.2. Powerplant description

The first component considered is the piping, and its preliminary straight path is shown with a blue line in Figure 2. As can be seen from the figure, the area is steep and thus it would require significant effort to transport and instal the pipes. However, there is already know-how regarding such installation in the region, which would make this procedure easier. The resulting pipe length reached 1500 m using a 500 mm diameter to obtain an efficiency of 95.4% when operating the turbine at maximal discharge, i.e. 0.71 m³ s⁻¹.

Regarding the power generation, a Pelton turbine of 4.2 MW would be used. With a rotation of 1500 rpm, about three injectors are needed using a vertical axis (Zhang, 2016). The advantage of the vertical axis is to allow the distribution of the injectors to obtain a quasi-zero radial force on the turbine shaft. However, maintenance is a bit more complicated.

To pump the water back, several scenarios should be considered and will depend on the owner needs. If there is a need to pump back at the same generating discharge, two pumps of minimum 3.4 MW each will be required. An alternative scenario would be to lower the pumping discharge to 0.5 m³ s⁻¹, which would require a single pump of minimum 4.3 MW. This scenario is chosen in the following description. To ensure that the pump operates cavitation free it will be located several metres below the minimum height of the lower reservoir, which has an impact on the powerhouse construction. Therefore, the quaternary set configuration will be used.

A powerhouse needs to be built near the Fionnay reservoir to accommodate the machines along with their electronics. It is assumed that a building with a footprint of 150 m² is sufficient. However, as mentioned previously, this powerhouse will need to go several metres deeper than the minimum height of the lower reservoir to meet the requirement of the pump NPSHr (Net Positiv Suction Head required) and also have a part above the reservoir to evacuate the flow downstream the Pelton turbine. If it is impossible to build deep enough to instal the pump, another alternative is to build a second powerhouse located lower than the lower reservoir. As the lower reservoir is in a mountainous area this is feasible. However, this solution would require having both pumping and generating pipes and both powerhouses would need to be connected to the grid, which will be more costly. Assuming that a single powerhouse is required and can be built near the lower reservoir, the connection to the grid can be done via the neighbouring cable car. The distance between the two is approximately 220 m.

3.1.3. Investment cost

The estimated investment cost is shown in Table 1 and its distribution is shown in Figure 3. The total cost is decomposed into five main parts: reservoir cost, piping cost, civil engineering cost, hydraulic and electric costs, and the electrical connection to the grid. To these costs a percentage is taken for the study cost and miscellaneous cost. The total investment cost reached approximately 8.2 million CHF, which corresponds to approximately 1.9 CHF/W.

Figure 3. Pie chart showing the investment cost distribution for the first site. E&H stands for Electrical & Hydromechanical equipment.

Table 1. Global investment cost for the Louvie–Fionnay site.

Reservoirs	0
Piping	2,237,000
Civil engineering	2,020,000
H&E equipment	2,227,000
Electrical connection	126,000
Study	697,000
Miscellaneous	488,000
Total (CHF)	8,158,000
Total (CHF/W)	1.9

3.2. Site Moneyeu reservoir–Les Vernays pond

3.2.1. Site description

The second site is located near Bruson, also in the Wallis Canton, where two reservoirs already exist. The upper reservoir is an artificial hillside catchment reservoir built in 2013–2014 to supply the ski resort with artificial snow. This reservoir is located near Moneyeu and is at an altitude of 1940 m with a water capacity of 14,000 m³. To ensure the full capacity of the reservoir before the ski season, this reservoir is connected by a 200 mm pipe to a pumping station (elevation: 1650 m, location: St.-Tarpe) with a pumping capacity of 80 L/s. The lower reservoir, called “La Gouille des Vernays”, is located near Le Châble. Its elevation is 792 m and it has an approximate water capacity of 170,000 m³. It is assumed, at this phase of the analysis, that a volume of water of 15,000 m³ can be used even though no interaction with the local industry using the lower reservoir has been carried out. The capacity for energy storage of this site reaches 47.0 MWh, providing a generating power of approximately 8.3 MW based on a cycle of approximately 5 hours in generating mode. Regarding access to the construction site, no particular difficulty is expected. The upper reservoir is located in the subalpine region where there are already roads that would allow trucks to easily reach the construction site.

3.2.2. Powerplant description

The preliminary piping path is shown with a blue line in Figure 4 and can be decomposed into two sections. The first section is located on the ski resort and the second section in the urban area. In the first section, the piping path follows first the path of a chair lift and second of a cable car. On the second section, the pipe path will go through a field and along a road to reach the lower reservoir. The total length of the path is approximately 4600 m. To ensure about 95% efficiency the pipe diameter is 600 mm, resulting in a head loss of 64 m at the considered generating discharge, i.e. 0.89 m³ s⁻¹.

Figure 4. Geographical location of the Moneyeu–Vernays reservoirs (red surfaces). The preliminary piping path connecting the two reservoirs is shown with a blue line. Source: Swisstopo.

The difference in altitude between the two reservoirs is very high, i.e. 1148 m, which is comparable to the height of the most powerful hydro powerplant in France: Grand Maison powerplant with a height of 955 m. As with the previous site, it is assumed that the know-how exists in the region to properly design and build this penstock and that the low discharge will ease such construction. At the stage of the preliminary study, only a complexity coefficient was used to reflect the required work in the investment cost. For such a high head, the turbine type selected is a Pelton turbine rotating at 1500 rpm with two to three injectors. If two injectors are chosen a horizontal axis can be selected; otherwise, a vertical axis is preferred.

The height of the site makes it challenging to obtain a pumping discharge similar to the generating one. According to the portfolio of a pump manufacturer, it is feasible to pump water at such a height but at a lower discharge. To limit the number of pumps to two, the pumping discharge is lowered by 50% in comparison with the generating discharge, resulting in a pumping discharge of 0.42 m³ s⁻¹.

Similar to the previous site, the powerhouse will need to reach sufficiently deep to allow the installation of cavitation free pumps. Also, the use of a Pelton turbine requires it to be installed above the maximum height of the lower reservoir. Therefore, a quaternary set configuration will be used.

For this site, a powerhouse needs to be built. It is estimated that a footprint of 150 m² is sufficient to accommodate all the machines and their electrical equipment. The powerhouse can be connected to the grid via the neighbouring industry. The distance needed is less than 250 m.

3.2.3. Investment cost

The estimated investment cost is shown in Table 2, and its distribution is shown in Figure 5. The total investment cost reached approximately 15.5 million CHF, which corresponds to approximately 1.9 CHF/W. This price per watt is closed to the one obtained with the first investigated site.

Figure 5. Pie chart showing the investment cost distribution for the second site. E&H stands for Electrical & Hydromechanical equipment.

Table 2. Global investment cost for the Moneyeu–Vernays site.

Reservoirs	0
Piping	4,556,000
Civil engineering	2,032,000
H&E equipment	5,299,000
Electrical connection	482,000
Study	1,322,000
Miscellaneous	926,000
Total (CHF)	15,471,000
Total (CHF/W)	1.9

3.3. Comparison with large hydro projects

3.3.1. The Limmern pumped storage powerplant

The Limmern pumped storage powerplant (LPSP) is one of Axpo’s most important expansion projects in recent years, with investments amounting to CHF 2.1 billion (Axpo, 2020). LPSP was commissioned in 2016/2017 after a 10-year construction and planning period. It is located in the Glaris canton and features two artificial reservoirs. The upper one, called Muttsee dam, is at an altitude of 2474 m and can contain up to 23 million m³ of water. The lower one, called Limmernsee dam, is at an altitude of 1885 and can contain up to 92 million m³ of water. The expansion project increased the output power from approximately 520 MW to 1520 MW with four GE 250 MW variable speed pump-turbines (Kellner, 2016). Therefore, the expansion cost is about 2.1 CHF/W.

3.3.2. Nant de Drance pumped storage powerplant

The Nant de Drance project started in 2008 and its commissioning was fully completed in June 2022 (Nant de Drance, 2020). It consisted in the construction of a new underground powerplant in the Wallis canton. It features six GE 150 MW variable-speed pump-turbines (Beller, 2020) and connects two existing reservoirs: Émosson dam and Vieux-Émosson dam. The Émosson dam is at an altitude of 1931 m and can hold up to 225 million m³ of water. The height of the Vieux-Émosson dam was increased by 20 m to double its water capacity. Its altitude is 2225 m and it can hold up to 25 million m³ of water. This new project cost about CHF 2.1 billion (ATS, 2019), i.e. 2.3 CHF/W.

3.3.3. FMHL+

The FMHL+ project consisted in an expansion of the former Forces Motrices Hongrin-Léman SA (FMHL) pumped-storage powerplant. The first powerplant was commissioned in 1971; it connects the Léman lake (372 m) to the Hongrin dam located in the Vaud canton at an altitude of 1255 m and can accumulate up to 52 million m³ of water. The FMHL powerplant accommodates four horizontal ternary sets each composed of one pump and a turbine composed of two Pelton wheels for a total installed capacity of 240 MW. The expansion, commissioned in 2017, consisted of building a new powerhouse cavern and connecting it to the existing head race and tail race. It accommodates two vertical ternary sets composed of one pump and one Pelton turbine of 120 MW (Rouge et al., 2016, 2017). Therefore, the extension allowed a doubling of the installation capacity for a cost of CHF 0.331 billion (Alpiq, 2017), i.e. 2.8 CHF/W.

4. Conclusion

An initial study showed locations where natural or artificial reservoirs are located in the French-speaking part of Switzerland. From these, two promising sites were selected, and a pre-sizing was carried out regarding building new small pumped-storage powerplants, along with their investment cost.

The first site connects Louvie lake to the Fionnay reservoir and has a usable water volume of approximately 17,000 m³, with a head reaching 722 m. The energy storage capacity for this site reaches 33.4 MWh, providing a hydraulic power of 5 MW based on a cycle of approximately 6 hours in generating mode. To fulfil this requirement, one pump and one turbine in a quaternary set configuration are selected. The preliminary investment cost reached 8.2 million CHF, corresponding to 1.9 CHF/W.

The second site connects the Moneyeu reservoir to the Vernays pond. The volume of usable water is approximately 15,000 m³ and the head reached 1148 m. The energy storage capacity for this site reaches 47 MWh, providing a generating power of approximately 8.3 MW based on a cycle of less than 5 hours in generating mode. Because of the high head, the pumping discharge is reduced by 50% in order to limit the installation to using only two pumps. The resulting installation is composed of a quaternary set type which connects the motor to its pump. The estimated investment cost reached 15.5 million CHF, corresponding to 1.9 CHF/W.

In the present analysis, the estimated cost of the two small hydro projects is lower than that of recent large hydro projects. One reason is that the cost was compared against built price, which is almost always higher than the initial expected cost. Another reason is that only sites with existing reservoirs were considered. Furthermore, these large projects implied, for example, the construction of cavern powerhouses or the heightening of dams, which are much more complex than the construction considered in the present analysis. As an order of magnitude, it can be estimated that the cost of reservoirs represents 33% of the total cost. For this reason, the evaluation of the potential of small PSP (Pumped-Storage hydropower Plants) was focused on sites with one or two existing reservoirs (Figure 1) with the required water volume. It is not surprising that the two most promising sites are sites with two existing reservoirs in close proximity to each other, as well as high heads, which reduces the size and cost of the hydraulic and electromechanical equipment. However, high head requires high-performance pumps, especially in terms of cavitation. It should also be noted that the cost estimation is based on the literature or quotes obtained from suppliers. It does not take into account the constructive specificities of each site. The needs of the pump NPSHr requires the pump room to be implanted at a great depth, the cost of which is strongly dependent on the geological nature of the soil and the methods of construction used. In addition, the present software is focused mostly on the technical part, whereas some additional costs should be included, such as the administrative cost. An example of administrative factors would be carrying out a study assessing the environmental impact of the project or dealing with objections against the project, which could significantly impact the cost.

Finally, the cost shown in the analysis is an estimation to provide future customers with the order of magnitude of the investment cost, which is crucial to validate the choice of site and allow customers to pre-evaluate the types of services they could offer, such as energy storage or frequency containment reserve. Last but not least, the presented costs should be, once again, taken cautiously. This is a preliminary study. A margin of approximately 25% should at least be taken into account. The reason can be illustrated with the estimation of the penstock cost. In this study, the order of magnitude of the cost was discussed with people in the field. But to obtain a more accurate cost, several studies are required. One of them is a geological study with the required civil engineering works. The results of such a study can change significantly from one place to another.

Acknowledgements

The authors thank the SCCER FURIE and ALTIS for their financial support.

Disclosure statement

No potential conflict of interest was reported by the authors.

Data availability statement

The data that support the findings of this study are available from the corresponding author, OP, upon reasonable request.