How Hybrid PV Technologies Can Contribute to the Decarbonisation of Thailand's Power System

While Thailand’s power generation is currently characterised by a high share of fossil fuels (81% of total electricity generation in 2021 came from gas and coal), the country has tremendous solar PV potential, both at utility scale and for rooftop PV, thanks to high irradiance and high daily solar exposure.

Solar PV is a low-cost source of clean electricity with high potential for decarbonising power systems. It is, however, a variable resource, as generation fluctuates with the weather.

In systems with growing shares of variable renewable energy (VRE) sources, power system flexibility plays an increasing role, as can be seen in the IEA NZE Scenario, in which global electricity system flexibility quadruples by 2030.

As flexibility needs can differ depending on the timescale, situation and share of VRE, the IEA has developed a framework of six phases to assist policy makers in identifying the needs related to system integration of VRE.

The classification into phases is key in order to identify the flexibility needs of the Thai power system both now and as the system evolves towards further decarbonisation. Currently, the Thai power system is in Phase 1, transitioning towards Phase 2, which means that system operators are starting to notice VRE generation. In Phase 2, short-term flexibility needs will be the most crucial to address. Increased medium-term flexibility will also be key as the system moves towards higher shares of VRE, as well as in Phase 3.

Integrating growing shares of VRE into power systems requires a shift in the way the system is operated, as the determining factor for operation becomes the net load, meaning the difference between the system load and the output from VRE generators. When weather conditions change, there can be strong ramps in net load as the output of VRE generators changes. Each technology’s benefits and drawbacks must be considered when creating decarbonised power systems. It is also important to consider how the interplay between these different technologies can bring additional value to the system.

We assess here the ways that selected clean technology options – solar PV, battery energy storage systems (BESS), hydropower and hybrid PV – add value to power system operations, and how they can be utilised in an optimal way.

Variable renewable energy sources such as solar PV can provide flexibility if their operation plan allows it – and, when they are viewed over a large region, their aggregate profile is less variable. For example, a study that considered several solar PV plants located over a radius of 100 km showed that the production variability was lower than when looking at each single plant’s output. Similarly, the ramp rates of solar PV production were de-correlated in distances over three km in timescales under five minutes.

While the installation of solar PV on water surfaces is a relatively new practice, the interest in it has increased in recent years. The main advantages of floating solar PV are that it avoids competing for land use, it increases the yield provided due to the cooling effect of the water, and it has the potential to reduce both water evaporation and water temperatures. However, the installation of solar PV on water surfaces is more complex than on land. Since the panel anchoring cannot be subjected to high variations in water levels, currents or waves, it could introduce additional locational and operational complexities.

BESS offer fast and accurate response times to dispatch signals from system operators. Their modularity enables a tailored deployment, and they are able to provide multiple power system services such as energy and ancillary services, which contribute to the fulfilment of reserve requirements and reduce congestion.

Hydropower is the largest renewable electricity source globally. It can offer a cost-effective solution by contributing to flexibility, energy storage and ancillary services. It can also provide several different flexibility timescales and will be especially relevant for power systems in Phase 3 or higher of VRE integration, in which the ability of hydropower to dispatch and quickly ramp up can address changes in net load.

A common misconception is that in order to harness these complementarities between technologies, a hybrid plant setup is needed.

Hybrid generation technologies combine different energy generation, storage and conversion technologies. The way the different constituents are combined, both in terms of location and operation, will define their degree of hybridisation and how they are perceived by system operators. It is therefore important to distinguish between the levels of hybridisation, which we do following the taxonomy of the Renewable Energy Laboratory (NREL) at the United States Department of Energy:

• Co-location, where a geographical link exists between constituents (through a shared point of interconnection for example), but where the operation of the different plants is independent. From the perspective of grid operation, co-located resources do not differ from two individual technologies as they are controlled individually.
• Virtual power plants, which do not have a geographical link, but an operational one, which ensures a coordinated control of the technologies, taking advantage of their synergies.
• Full hybridisation, where both a locational and an operational link exist, and components are often shared. Here, the technologies are integrated fully so that at the point of interconnection, the hybrid resource is treated as a single resource and from a system operator’s perspective, the constituents are therefore seen as a single plant.

Leveraging the complementarities between technologies is, however, possible as soon as the different technologies are integrated into the same power system.

To illustrate this, we analysed the effect of adding 5 GW of solar PV to the power system on the hydropower generation profile using the production model from the IEA Thailand Power System Flexibility Study. Three scenarios were considered, each at horizon 2030 and each considering the accelerated deployment of VRE based on targets from the ASEAN Interconnection Masterplan Study (AIMS-III) project (15% share):

• a base case 1
• the “Additional Solar Scenario” in which the 5 GW of solar PV are added at an unrestricted location close to the load centre (in this case, the Metropolitan Bangkok (MAC) region)
• the “Additional Solar, Co-located Scenario” in which the 5 GW of solar PV are co-located with hydropower reservoirs in the North-eastern (NEC) and Northern (NAC) regions, and the output of the co-located plants cannot exceed the capacity of the hydropower plants (assuming shared infrastructure).

By comparing the base case with the two “additional solar” cases, it is apparent that the addition of solar PV to the power system causes the hydropower profile to adjust its daily generation profile to accommodate the production profile of the solar resource. Hence, in both “additional solar” cases, on average there is higher hydropower generation in the night hours than in the base case – while during the day, hydropower generation is lower.

This demonstrates that hydropower provides intra-daily flexibility to the system as the share of solar PV increases independently from co-locating the two resources.

While hybridisation has many advantages, it also adds locational and operational constraints, as well as a loss of visibility to system operators. These disadvantages need to be assessed and weighed against the benefits in order to ensure that the plant’s construction provides net benefits to the system.

First, in co-located or fully hybridised setups, the developer constructs multiple installations (e.g. solar, wind, battery or hydropower) in the same location. The chosen location, however, may not be ideal for any one of them, which would introduce the need for a compromise when co-optimising the location. This could reduce the potential of individual constituent technologies to provide system services such as non-wire alternatives, or reduce the potential yield.

For example, by installing solar PV and BESS in a hybrid setup, the location decision will be the result of a compromise between the optimum solar resource and the potential of the battery to relieve congestion. It is possible that installing the two resources in different locations could produce more solar PV electricity (due to higher solar irradiation at that location for example), while allowing the battery to serve as a non-wire alternative and thus reduce congestion.

This trade-off is particularly clear when considering hybrid hydropower projects. As the locations suitable for hydropower plants are extremely limited, from the perspective of solar PV siting, the joint locations are often suboptimal.

The decision in favour of hybridisation should therefore be based on verification that the system-level costs of the locational drawbacks are offset by the advantages of the hybridisation. Where in certain cases, the cost-benefit analysis might be positive for the plant operator, it is crucial to check that the setup makes sense from the system-level perspective.

Secondly, hybridisation can also add operational constraints, which result from the shared infrastructure or the operational linkage. In terms of infrastructure, and in the case of a retrofit addition of solar PV to existing plants, a constraint may exist in terms of the capacity of the connection point (described below). Furthermore, in the hybrid PV and BESS setup, sharing an inverter, while bringing some advantages as outlined above, limits the output in times of high generation. In the case of hydropower, the operation range of the plant could limit the full dispatchability of the hybrid system. Hydropower ramping for short-term flexibility, for example, could be constrained by the allowed volumes of water discharge downstream.

Coming back to our modelled example, and looking at the differences between the two “additional solar” scenarios, one can see that there are slight differences in the way the hydropower generation is adjusting over time. In the co-located scenario, to take into account the shared infrastructure of the hydropower/floating solar plant and the constraint on the PV generation due to the reservoir’s location, two additional constraints were introduced into the model. Firstly, the scenario is modelled assuming the use of common transmission and distribution equipment and is therefore limited to the capacity of the existing hydropower plant (5.34 GW) at which they are located. Secondly, there is limited transmission capacity between the hydropower plant that is in the northern region of Thailand and the main load centres located in the centre of the country. These two constraints lead to slight differences between the two “additional solar” cases. Indeed, the adjustment of the hydropower profile to solar PV availability is stronger in the co-located scenario, as the capacity of the shared infrastructure limits the total output at the site. The fact that the hydropower resource is increasingly used to balance local congestion instead of being controlled at system level reduces the cost-efficiency of the system.

Lastly, in full hybrid or virtual power plant setups, a trade-off is the loss of visibility from the system operator’s perspective. Indeed, the fact that the hybrid technologies’ constituents are controlled together renders their individual behaviours invisible to grid operators, who then rely on plant operators to ensure operations are managed in a system-optimum way. Providing incentives and regulations to plant operators to ensure that operations are system-friendly can address this drawback.