Hydropower is one of the most reliable, sustainable, renewable, and cheapest sources of energy [1
]. In a climate change scenario, hydropower projects are also an adaptation measure for the use of water resources, since regulated basins with large reservoir capacities are more resilient to changes in these resources [2
]. However, small-scale hydropower plants are mostly run-of-river (RoR) systems, located in mountainous areas, without dams or significant water storage. Despite being one of the most profitable energy technologies, with relatively low operating and maintenance costs, and being environmentally benign [3
], this type of facility has the disadvantage of having an irregular production that is subject to RoR flow, which in turn is highly variable depending on the duration of rainfall and snow cover. Since RoR plants usually have a small or non-existent storage facility that allows for very short-term water storage, the hydropower plant does not have enough water to continue operating when the inflow falls below the turbine’s minimum discharge value. Another drawback of these systems is that when the inflows are extremely high and the turbine capacity is exceeded, the water will be wasted (spill) and will become a lost opportunity for generation [4
Europe is the market leader in small-scale hydropower technology, with Spain, Italy, France, Germany, and Sweden being the main producers [5
]. However, the potential of RoR plants has not yet been fully explored and exploited, so there is interesting scope for the development of this technology and its optimization [5
]. Management has to be done with some particular plant operating conditions but also with some environmental requirements. In this context, decision makers need to have information on the flow in the river in the short and medium term. If they could forecast the energy that will be produced in the following months, they could better adjust market prices and plan maintenance or other energy resources when the water availability falls below the minimum operating flow.
Although some forecasting models already exist in the area of small-scale hydropower production [6
], there is still a gap in linking forecasting with decision support systems. In most cases, water system managers usually make decisions based on historical statistical data despite the great advances in climate forecasting that exist so far. There are currently a number of European projects (e.g., MED-GOLD, CLARA, S2S4E, H2020_Insurance, VISCA, CLIMATE-FIT, SECLI-FIRM, PROSNOW, CLARITY) involved in the development of tailored and usable climate services (CSs) that can facilitate the uptake and use of climate information and forecast by the final users. In the framework of the H2020 project CLARA (climate forecast-enabled knowledge services), a CS targeted at end-users and able to support small hydropower systems’ management, has been developed. A preliminary version of this tool is presented in [10
], showing the structure of the tool and utilities but without delving into details, such as the data and models used and the co-development process outcomes. This preliminary version used a single forecast. However, as probabilistic forecasts are based on ensembles, in this paper, we used a set of forecasts that represent a range of future weather possibilities. Multiple simulations were run, each with a slight disturbance of the initial conditions of the weather models [11
]. The new version presented here includes the range of probabilistic forecasts in order to show end-users the range of weather possibilities. The aim of this work was to also show the main outcomes of the co-generation process in order to provide a short guide to help future CS developers to connect climate forecast data with the decision support process. It thus provides very useful knowledge for technicians in charge of control operation centers of small hydropower plants and managers in regional administrations.
The tool proposed in this paper is a state-of-the-art CS based on the newly created Copernicus Climate Change Service (C3S) [12
], which provides Earth observation satellite-based information and in situ (non-spatial) data. The climate data developed by C3S, in this case SEAS5 [13
], provides seasonal forecast data as input to the developed CS. A strong point of this CS is the cogeneration methodology used during its creation and development. In this way, the proposed service bridges the gap between data providers generating climate impact data, on the one hand, and administrators and policy makers, on the other. In addition, the service is based on databases and models created with a scalable architecture that allows their application in other systems. This ensures that the information available is useful for hydropower management at different scales, both at the local and regional level across Europe, which contributes to the marketability of this type of information. More generally, this paper aimed to highlight the value that CS tools can bring to the hydropower energy field but also to highlight relevant outcomes about the skill of the forecast data when applied at the local scale in a pilot area in southern Spain.
The paper is organized as follows: Section 2
introduces the methodology carried out for the service development, including a pilot area description, the data and models used, and the CS approach, which also illustrates the CS workflow; Section 3
presents the main outcomes of the co-design process, the final structure of the web user interface, and service output examples; and Section 4
contains the findings and implications of the main results.
2. Materials and Methods
2.1. The Pilot Area in Southern Spain
The CS was tested in southern Spain, in a Mediterranean high mountain area where snow has a critical influence over the hydrology of the downstream areas. The RoR pilot system consists of three small consecutive hydroelectric plants in the Poqueira River basin (Figure 1
) belonging to a leading company in the Spanish energy sector, with a combined generation capacity of between 10 and 12 megawatts. The study area is divided into three catchments of interest for the user. These three catchments define three different points of water uptake for the hydropower production. Their main characteristics are compiled in Table 1
The basin is located within the Sierra Nevada Mountain Range, which is a national park and biosphere reserve. This explains the special importance of carrying out adequate management of water resources in strict compliance with environmental regulations. It is an alpine/Mediterranean climate region with a highly variable rainfall regime. Annual cumulative values range from 1000 mm in wet years to 200 mm in dry years [14
]. Snow appears recurrently at altitudes above 1000 m a.s.l. and is more persistent at altitudes above 2500 m a.s.l. from November to May. Snow cover is subject to several accumulation-ablation cycles during the snow season [15
]. The average annual area of fractional snow cover was 0.21 m2
between 2000 and 2013, and ranged from 0.9 to 0.16 m2
in wet/cold and dry/hot years, respectively, with an average standard deviation of 0.23 m2
]. The spatial distribution of snow cover is very heterogeneous over the years and very difficult to predict. Therefore, this pilot area was a perfect candidate to apply and analyze the potential of the proposed CS.
Under these circumstances, climate is extremely variable and seasonal forecasts usually show very limited skills and performance. However, the use of a seasonal forecast in this pilot area allows managers to estimate the production in the next 7 months thanks to the knowledge of the water availability in terms of the volume of snow (from the real-time hydrological state of the contributing basin to the RoR plants) together with information of the seasonal forecast of river inflow.
2.2. Data and Models
Several data sources were used in this work: Historical information, seasonal forecast data, and local data, with all of them provided in the collaborative framework of the CLARA project.
On the one side, the service includes dynamic data, which is seasonal forecast data and simulated hydrological data, that can be updated as better knowledge of the physical environment and as more measurements become available, resulting in improved data forcing and models:
Seasonal (which go up to a seven-month prediction lead time) forecast of daily river flow data issued monthly by the Swedish Meteorological and Hydrological Institute (SMHI). SMHI produces these data by forcing the E-HYPE (European Hydrological Predictions for the Environment) model with the ECMWF SEAS5 seasonal forecast. SEAS5 is based on a global climate model, which, since the oceanic circulation is a major source of predictability in the seasonal scale, is based on coupled ocean-atmosphere integrations [17
]. E-HYPE is the European setup of the HYPE model, which calculates hydrological variables on a daily time step at an average sub-basin resolution of 120 km2
over the entire continent [17
]. Figure 1
shows the location of the sub-basin where seasonal forecast data are produced, which has a size of 527 km2
. Probabilistic forecasts are produced as an ensemble of scenarios that present the range of future river flow possibilities. In the service testing stage, we used the SEAS5 hindcast period 1981–2015 for each calendar month and up to seven months ahead, considering an ensemble of 51 members. In this work, the raw seasonal forecast data were presented at a monthly scale and downscaled to the intake points of the three RoR systems to match the temporal and spatial scale suitable for this particular application. This was done by using a quantile mapping methodology [20
], usually adopted as a bias correction method, which leads to a good performance [21
Interpolation of the meteorological real-time data and current state of the hydrological variables were extracted from GMS-Snowmed service [23
], which makes use of WiMMed (Water Integrated Management in MEDiterranean Environments) [24
], a physically based and fully distributed hydrological model. This service makes use of past and quasi-real-time observations of daily hydro-meteorological data (precipitation, temperature, river flow), from different meteo-hydrological networks in the area (Red Guadalfeo, SAIH Guadalquivir, RIA-JA, Red Hidrosur). The outputs of GMS-Snowmed directly offer distributed information about the antecedent weather and current water availability in the basin upstream from the RoR plants at daily and monthly scales.
On the other side, the service makes use of static data related to local specific facility features defined by end-users:
Past observations of daily streamflow measurements provided by the managers of the hydropower system and available for the period 1969–2018. These data provide a very adequate overview of the historical river inflow to the RoR system.
Some records related to the specific consumption of the turbines that are also provided by the managers of the hydropower system. This information is mainly used to compute the production of the hydropower plants.
Threshold value of the target indicators in the service, according to the turbine’s minimum and maximum discharge, provided by the managers of the hydropower system.
Minimum environmental flow restrictions, as defined in the Hydrological Plan of the Mediterranean River Basin, the water authority in the study site.
Different scale issues were considered in the development of the service to achieve a satisfactory representation of the flow data on both the historical and forecasted information. On the one hand, there is the more detailed scale of the data and models, and on the other, the broader scale of the industrial processes that determine energy production. At the core of the CS, the calculation scales of the historical and the real-time modules are determined by the characteristic scales of the physical processes. The stream discharge is determined by the combined action of snowmelt, surface runoff, infiltration, and subsurface and groundwater flows. WiMMed performs hourly calculations of the energy and water balance on a 30-m gridded representation of the terrain, providing input data to circulate both surface and sub-surface flows throughout the catchment area to the selected outlets. Despite the user’s requirements for decision-making being framed in a daily time step, which is the data scale displayed in the service, the hourly calculations are accumulated or averaged to the daily scale at sub-basin level (as represented in Figure 1
). This avoids a non-adequate performance of the model when the non-linear effects of the topography of the area and the significant scales of the energy fluxes’ evolution cannot be neglected. In addition, the results are also aggregated on a monthly scale to provide a perspective within the current year compared to previous ones. The historical simulation of the river flow generated by the model results in a correlation coefficient r of 0.790 and 0.864 on a daily and monthly basis, respectively. As for the final downscaled seasonal forecast, which is not modeled in as much detail as the historical one, it is displayed on a monthly basis.
2.3. Climate Service Approach
The CS was developed following a co-generation process, where data providers and service purveyors and also end-users are closely involved in the design of the tool and local data provision (Figure 2
). Co-generation entails that all participants are engaged on an equal footing in an effort for co-designing, co-developing, co-delivering, and co-evaluating CS tools [27
]. The outcomes and conclusions of this co-development process are presented in Section 3
shows an overview of the structure of the service, called SHYMAT (Small Hydropower Management Assessment Tool), including the inputs from different data sources, the data processing workflow, and the service outputs.
Firstly, the historical and quasi-real-time hydro-meteorological information was collected from the GMS-Snowmed service, and the seasonal forecast of daily river flow. River flow was the variable selected as an indicator of the available water to generate electricity. With the aim of making the downscaling generation of the river inflow, river flow forecast data were combined with historical local measurements of the river flow in the RoR system’s uptake point. Moreover, a set of local specifications, such as historical energy production, turbine maximum and minimum discharge, turbine-specific consumption, and environmental flow requirements, were defined by end-users as well as the outputs answering their needs.
Once all the seasonal forecast, basin and river data, and local facility data were collected, a data model defining the different components of the hydropower system and including all available information was developed. The CS uses the data model in order to automatically build a scalable database and a scalable topological scheme of the pilot RoR system. Both scalable capacities allow the implementation of SHYMAT in other sites. Thus, the on-line implementation of the CS included two main aspects:
A web administration panel: A CS administrator (or also a CS developer) accesses a panel in order to define, store, and manage the information to be included in SHYMAT. Through this administration panel, a CS administrator is able to create users, hydropower systems, the elements that compose the hydropower systems (rivers, load chambers, hydropower plants, basins, power grid), and also import the available data (related to climate, hydrology, energy production, and local facility data).
A web-user interface: A CS user accesses a graphical application, which allows fast and intuitive access to all the information included in SHYMAT. This user interface has capacities for SIG geolocalization, user registration, data processing and acquisition, and graphical monitoring and supervision of the hydropower systems. It includes two different modes: One for “historical information” and another one for “forecast information”. The historical mode is devoted to showing weather and hydrological past and real-time data. The forecast mode includes hydro-meteorological forecast information and offers the user seasonal forecast information together with a prediction of the operability of the plant and the energy production expected for the next seven months (because, as it was already said before, seasonal forecast data go up to a seven-month prediction lead time).
The implementation of SHYMAT was carried out following a view controller model (VCM) methodology. One of the main advantages of the VCM architecture is that it allows for fast and collaborative development but also an easier update of the application. This characteristic is very relevant for the CS development, because SHYMAT is expected to be updated as the knowledge of forecast information progresses. In addition, the VCM architecture can assist decision-makers and help practitioners to prepare and realize such integration projects [28
]. The view is the user interface on which the end-user can interact and perform some actions. The controller is the part of the application in which data are processed after a request is received from the view and before updating anything in the database with the model. When a CS end-user sends a query to the view (web user interface), the controller (web administration panel) asks the model for information from the database, which answers the controller by sending the requested information. As a reply to this request, the controller sends the information to the view, which automatically collects the updated information and shows the required data on the screen.
4. Discussion and Conclusions
The operation planning of RoR plants should not be solely based on historical local data, since water availability presents a very high interannual variability, which is even more significant in mountainous Mediterranean areas, where snow cover and snow processes have a large influence on the quantity and timing of water availability. Thus, seasonal forecasts constitute an added source of information that may help to narrow down the operational options inferred from historical data sources. The proposed CS provides end-users with the most up-to-date hydrological data, combining measurements and modelling with the most advanced seasonal forecast that currently exists at the European level. The co-design process gives as a result a tool that perfectly matches the user’s needs, i.e., the information uses the correct scale and the right tools to convey information, which results in a more effective knowledge system [29
]. The service provides simple and intuitive results, while being supported by the state-of-the-art knowledge on snow hydrology and weather forecasting.
SHYMAT shows hydropower managers how seasonal climate forecasts can provide advanced information about water availability for the next season, and users can make use of this knowledge for several aspects related to plant operation planning. Users can take advantage of the climate forecast in order to anticipate: (1) Periods in which there will be enough water to turbinate (periods with production); and, on the other hand, (2) periods in which there will be an inadequate amount of water to turbinate, and maintenance tasks can be planned (periods without production). Moreover, having knowledge of the possible water excess discharges coming from snowmelt, which may result in the spilling of water, gives managers the opportunity to quickly tune up additional turbines. Finally, great value comes from the prediction of energy production, which is clearly valuable information for market issues. In addition, the development of this kind of tool, addressed to hydropower managers to predict the operability of the plant and the expected energy production, should consider not only forecast information but also past data at the local scale. Both types of information provide end-users with more reliability and trust.
SHYMAT is a scalable solution, as service developers just need to define the elements of the new hydropower system and a new topology scheme will be automatically generated to include the new local information and parameters. This allows an easy application in other systems, with very low development costs, which helps to bring the C3S information to other sites in Spain and Europe, while also contributing to the bloom off climate services as an emerging market. In addition, the CS combines service-oriented architecture (SOA) and VCM architecture for cloud services and information systems, as other works have also proposed [28
The findings presented in this work have enormous implications in the emerging markets of climates services, helping future climate service developers to connect climate forecast data with the decision support process in the hydropower sector. As previous works have indicated [29
], technical information, such as historical and local data of water availability and production, placed in the context of hydropower plant operation according to the management experience, leads to more robust knowledge and contextual applicability of the seasonal climate forecast. This will also help end-users to overcome some perceived barriers, such as the accessibility, relevance, and usability of seasonal climate forecasting [30
Keeping these general ideas in mind, the major problem to solve is how to generate forecast data with the highest skill and reliability possible. In this work, a quantile mapping approach was adopted as a direct and easy-to-run method to downscale the raw forecast data to the local spatial resolution required by the RoR system; more complex methodologies for stochastic bias correction [31
] can be applied to further overcome these scale issues. Moreover, future research could focus on the improvement of forecast information at the local scale by using both local historical data and high-resolution model outputs with better performance when reproducing the local results.
The energy demand was not mentioned in this work. One might think that periods with low energy demand may be less sensitive to uncertain forecasts and, on the contrary, for periods where the demand is high, the energy producers might be more sensitive to uncertainties. However, the pilot system is composed of small hydropower plants and as such is fully dependent on the seasonal availability of water, whose regime makes non-operational conditions usual during a significant fraction of the year; the operation criteria is to generate the potentially maximum amount of energy at every moment, and this is easily absorbed by the energy market at any time given its magnitude. In such systems, the key advantages provided by seasonal forecasting is to optimize the operation of the facilities, mainly the seasonal start up and closure of the system, together with the dynamic seasonal planning of generation and cost-benefit analysis. Demand dynamics is key in hydropower systems with a storage capacity, and the requirements that its inclusion involves are needed in the future for the extension of SHYMAT to larger hydropower systems, with greater variability of both the demand and the response capacity.
Finally, it must be highlighted how the co-generation approach followed to develop this tool has benefited the result. As the tool evolved, the engagement of the end-users in the pilot case was increasingly enthusiastic, mostly due to the natural uptake of comprehension of both the software but also the underlying know-how. Despite being aware of the limitations forecast data on a seasonal basis still show, the joint discussion of needs for downscaling, and the available options, and requirements of the information to be provided by the service resulted in an efficient exchange and integrated acquisition of expertise for both sides, end-users and developers. End-users valued the possibility of a seasonal forecast that also includes an assessment of skill and uncertainty, leading us to shape their needs for information to understand and frame such additional metadata in the forecast, and to design optimal visualization tools for them to understand and further exploit the information from the service. As a consequence, end-users are excited about the chance to implement the operational version of the service in their system since they tested the pilot version and had the opportunity to compare their day-to-day decision-making with and without the forecast. Even though a higher skill is still needed to narrow the probability of underperformance of the seasonal forecasts that feed the service, the operability of this tool takes them one step further in the future improvement of the service and provides training for staff to cope with the challenging conditions that future climate scenarios provide to the energy and water sectors. The co-generation process has resulted in their convinced willingness to participate in similar experiences in the future, and incredible opportunities for this service as a market product in the near future in a co-participation of public research institutions and private spin-off companies.