Three aspects of balancing - interview with Ghazal Ghayour Vahdat
Ghazal Ghayour Vahdat is a PhD student in energy systems at Lund University, supervised by Olof Samuelsson. Ghazal's PhD is the Lund part of the project "Three aspects of balancing", which looks at how renewable generation, the grid, storage, and new consumption should be expanded together.
Ghazal, what is the main goal of the project?
"To determine how production, storage, grid, and new consumption should be developed together to reach a system with a high share of renewables, using whole-system optimization that provides yearly trajectories. In my part, the goal is to quantify the value of coordinating storage and transmission investments, and to identify which mix of flexibility sources, under which conditions, keeps grid expansion, load growth, and renewable growth manageable together, while keeping the results grounded in the real system through realistic data, European cost sources, and larger, more realistic networks.
My own question in the project is what is gained by planning storage and transmission together instead of separately, and under which conditions that coordination pays off, with fast renewable growth and rising load. I build optimization models, from DC optimal power flow to multi-year co-planning of storage and transmission and test them on a small test system before scaling up. I am now widening the work from batteries to other flexibility sources, for example, flexible demand, electrolysis, and hydropower, to see which combination keeps grid expansion, load growth, and renewable growth manageable together. To keep the results relevant to the real system, the models use realistic load and wind profiles, European cost data (ACER) and a Swedish value of lost load. The next step is to scale up to larger networks, a simplified Swedish corridor with N-1 security, and I would like to test the cost and network assumptions with the grid operators and industry partners in the project."
What do the three aspects of balancing mean, and why is it important to include all three in one project?
"The three aspects are generation, storage and new consumption. The project covers all three because they have to be balanced together, and each PhD student works on one part.
At Uppsala, Ibrahim Tuhin works on the generation side. Ibrahim models variable-speed pumped-storage hydropower and the fast frequency and voltage support it can give. At Chalmers, Sara Fogelström works on storage: which type, what size and where in the grid, so that it delivers what the local network needs. My work at Lund covers the system level. I plan how renewable generation, grid, storage and new consumption such as electrolysis should be expanded together over the years.
It matters to include all three in one project because they depend on each other, and the grid links them. Wind and solar are built quickly and new consumption arrives on its own schedule, while lines take years. A storage unit that is well designed and well placed locally still depends on whether the transmission network has room for the power it moves. Hydropower can give fast support, but its value depends on what the rest of the system needs from it. My planning model sets the system-level need, for example how much flexibility is required, when, and where the network binds. The other two parts show what each technology can actually deliver. If each part were studied alone, the answers would not fit together. My own results show this: the role the battery plays depends on which constraint binds in the network, not on the battery technology itself."
How "uncoordinated" is the system now?
"Largely sequential. Generation, grid reinforcement, and new demand are planned by different actors on different timelines. Renewables can be built in a few years, while new lines take 10 to 15 years. Wind and hydro are in the north, and the load is in the south. The effects are already visible at transmission level as congestion, curtailment, and large price differences. I do not have a single quantitative measure of this, so I describe it as a mismatch in time, in construction speed, and in place."
What system capacity enhancements do you think are possible?
Coordination. Deciding storage and line together gave the lowest cost in my study, 22 % lower than planning the line alone at 6 %/yr load growth.
Storage at the right bus. A battery at the wind bus absorbs curtailment, and one at the load bus serves demand locally. The model chooses between them.
Reinforcement. In my case the corridor goes from 125 MW to 225 MW. The line is built in every case where growth binds.
Timing. Storage can cover the years before a new line is in service, so the sequence of investments matters and not only their size.
Other flexibility sources. Flexible demand, electrolysis, hydropower and hybridised plants. I will add these to the same planning model to compare them with storage and with each other.
What feasible development trajectories do you see?
"I work with yearly trajectories rather than snapshots for single years. In my test system, wind grows 8 % per year and load 3 to 7 % per year. The trajectories fall into regimes. At low growth, the line alone is enough and no battery is selected. At medium growth, planning storage and line together lowers cost by a few percent. At high growth, the gain is large, up to 54 % of NPV at 7 %/yr, and co-planning is the only case with no unserved energy. In every regime the line is still built, so storage and line act as complements. The result is sensitive to battery cost, which I am testing next, together with uncertainty in wind, demand, and outages, and with other flexibility sources in the same model."
What are the overall main challenges?
Mismatch in time: energy is produced at different hours than it is needed.
Mismatch in construction speed: renewables are built quickly and lines slowly.
Mismatch in place: production in the north, demand in the south.
Credible cost data: current battery cost is from ACER for 4-hour systems only, and the cost decline comes from NREL (US) and is uncertain. I am looking for European sources for the decline and for other durations. (Australia has already started using 8h batteries at utility scale)
Model realism and size: a small test system without N-1 security, and long horizons that need representative days (four days stand in for 8,760 hours).
"Third source of flexibility": what does it mean? Is it an unexplored area?
"Flexibility traditionally comes from dispatchable generation and from demand that can shift. Storage is a third source because it can act on both sides: it is a load when charging and a generator when discharging. It moves energy through time, whereas a transmission line moves power through space at each instance.
It is not unexplored. Storage for balancing, sizing, and siting is well studied, and co-planning of storage with transmission and wind has been studied since around 2017 to 2020 (for example, Qiu et al. 2017 and Zhang et al. 2020, both cited in the project application). Less explored is joint long-term planning of all four parts (renewables, grid, storage, and new consumption) at yearly resolution, with realistic cost data, storage, other flexibility sources, and the grid decided together rather than one after the other."
Are there any results to present so far?
"The results so far are from a simplified test system, so I give them as approximate magnitudes."
Operation over 24 hours: an optimized battery needs roughly 30% less power capacity and about 15% less energy capacity than the fixed one, and still removes curtailment and load shedding.
Five-year planning: total cost with storage is roughly 60% lower than without it, and the load shedding in the last year is avoided.
Ten-year co-planning: at around 5% yearly load growth, planning storage and line together saves only a few percent compared with the line alone. At around 6% the saving is roughly a fifth. Only the co-planned case has no unserved energy.
Sensitivity: the value of coordination rises from a few percent of the total cost at lower growth to around half at higher growth. At the lowest growth rates no battery is selected.
Longer horizon: I have extended the model to 25 years with more realistic, sourced cost data. Whether a battery is built depends strongly on its cost, and I am testing this further.
What are the most important insights so far?
"The value of coordination depends on how fast load grows:"
Load growth | What happens | Gain compared with the line alone |
|---|---|---|
Low | The line alone is enough and no battery is needed | None |
Moderate (about 5% a year) | The gain comes only from postponing the line investment | A few percent |
High (about 6 to 7% a year) | Battery and line work together, and unserved energy is avoided | Roughly a fifth, up to about half at the upper end |
Storage and line are complements. A battery moves energy through time, and a line moves power through space, and the line is built wherever growth is high.
The network state sets the battery’s role. It absorbs curtailment at the wind bus, serves load at the load bus, or holds the corridor at its limit, depending on which constraint binds.
Battery cost matters. With more realistic costs, the result depends on how low battery costs fall, so cost sources are important to credibility.
What has the project revealed? Any surprises?
"The battery does not relieve the corridor; it holds it at its limit. With co-planning, the corridor stays at 100 % for two more years than with the line alone.
A battery built when needed beats a larger one built up front. In my runs, the optimized battery is smaller than the fixed one and still removes curtailment and load shedding, at lower cost. The model adds capacity year by year as load and wind grow, instead of building everything on day one. Storage alone never releases the corridor and still sheds load, so coordination with the line is what creates the value.
At low growth rates, no battery is selected at all."