Energy Storage, Renewable Energy, Solar

Finding the Energy Storage “Sweet Spot”

Issue 9 and Volume 22.

Key steps must be followed to find the optimum sized megawatt-scale Li-ion energy storage system for a large wind or solar plant.

By Michael Lippert, Saft

A megawatt-scale lithium-ion (Li-ion) energy storage system (ESS) can be vital in successful grid integration of a large wind or solar plant by addressing the intermittency and unpredictability inherent in renewable energy. The challenge, however, is sizing the ESS for maximum operational and financial benefit. This is because an ESS can have several distinct roles, and only by understanding its role and the specifics of its site can engineers specify the right ESS for the job.

Ramp Rate Control

Grid operators often must limit the rate of change at which power is injected into the grid-the ramp rate. The output of a photovoltaic (PV) array of several megawatts can drop by 70 to 80 percent in about a minute. The ESS, therefore, must discharge in a way that ramps the net facility output down smoothly over seven or eight minutes (Figure 1). The ESS can absorb or release energy when a sudden shift in wind or passing cloud causes a step change in output. Ramp rate control ensures that the facility ramps at a rate that is compatible with the power system. This is particularly true for island grids, because they lack the inertia of mainland networks and are susceptible to disruption, which could be caused by simultaneous uncontrolled ramping of several renewable facilities.

FIGURE 1: Comparison of PV generation (red) and ramp rate controlled output to the grid (blue)
FIGURE 1: Comparison of PV generation (red) and ramp rate controlled output to the grid (blue)

The ESS will experience many small charge and discharge cycles. Over the day, the cumulative energy charged and discharged in 24 hours, known as throughput, can amount to around two to three multiples of the capacity of the ESS (2C to 3C).

Typically, a 10 MW solar farm would be combined with an ESS capable of delivering 5 MW of power and storing 1.3 MWh of energy. The facility would operate at an average depth of discharge (DOD) of 6 percent and a cumulated daily energy throughput of 2.5 MWh, which is equivalent to 1.9 times the capacity (1.9C).

In contrast, wind generation generally varies at lower amplitudes so a typical 10 MW wind farm could be equipped with a 2.5 MW ESS, delivering 0.58 MWh energy storage. It would operate at an average DOD of 4 percent with a cumulated daily energy throughput of 1.9 MWh, or 3.2C.

Smoothing

Smoothing aims to keep production within a given forecast window. The ESS compensates for power sags and, like ramp rate control, it will experience many small to medium charge and discharge cycles with a cumulated energy throughput equivalent to several full charge and discharge cycles (Figure 2).

FIGURE 2: Smoothing of PV production in 30 minute forecast steps
FIGURE 2: Smoothing of PV production in 30 minute forecast steps

Power Shaping

Power shaping uses an ESS to shape the power output of a plant to deliver steady and predictable power like baseline generation (Figure 3).

FIGURE 3: Shaping ensures a stable power output with controlled ramping up and down in the morning and evening
FIGURE 3: Shaping ensures a stable power output with controlled ramping up and down in the morning and evening

An ESS used for a typical PV farm in this mode will deliver a large discharge in the morning, before charging up during peak daylight hours in the middle of the day and discharging again later in the day.

A typical example for shaping of a 10 MW solar power plant would be an ESS providing 5 MW power and 10 MWh energy. The average DOD would be 35 percent, with a daily energy throughput of 7 MWh, or 0.7C.

Peak shaving

Peak shaving is intended to reduce congestion on the grid at peak times (Figure 4). It is mostly used to reduce load in periods of high consumption. For example, the ESS discharges to supply consumption peaks, thus relieving the grid from supplying peak power.

FIGURE 4: Peak shaving can avoid the risk of curtailment
FIGURE 4: Peak shaving can avoid the risk of curtailment

The technique is also possible on the supply side. For example, the ESS charges (absorbs energy) when the PV or wind plant’s power exceeds a set limit. It releases energy into the grid later in the day once the peak has subsided. It thereby ensures that the output of a plant never goes beyond an agreed limit and avoids revenue loss through curtailment.

In both cases, peak shaving avoids or defers investments in grid infrastructure that would otherwise be necessary to cope with peaks in consumption or generation.

Frequency Regulation

Frequency regulation ensures grid stability by injecting or absorbing active power to keep the frequency inside its limits (Figure 5). In doing so, the ESS helps the grid accommodate more renewable sources.

FIGURE 5: Frequency regulation allows ESS to inject or absorb active power to or from the grid
FIGURE 5: Frequency regulation allows ESS to inject or absorb active power to or from the grid

The operational profile of the ESS will depend on the number and amplitude of grid frequency deviations. Typically, deviations are of short duration and only infrequently at full amplitude. Operational characteristics of the ESS can vary considerably depending on the specific application and location, especially the distribution of discharge cycles at each power level.

Frequently, operators request that the ESS associated with wind or solar plants also provide frequency regulation services on top of its initial smoothing or shaping function. Furthermore, fast reacting energy storage enables renewable plants to effectively contribute to frequency regulation services.

Developing an Energy Management System

An energy management system (EMS) must be developed to determine the optimum size for an ESS. An EMS considers inputs such as a site-specific PR or wind survey, knowledge of ESS performance and grid requirements (Figure 6).

FIGURE 6: Energy management system development requires several inputs
FIGURE 6: Energy management system development requires several inputs

Site specific inputs include grid code limitations and local legislation, as well as measured data on wind or solar power output. It’s important to use high resolution survey results from the actual site over a period of several months, ideally a full year. This accounts for local geography and variability to achieve accurate sizing of the ESS, which in turn achieves better financial performance.

The second set of inputs is the customer’s objectives for the plant’s power output. This is basically the mode of operation, which can include one or more of the roles previously explained and their precise technical parameters and limits, as well as economic variables.

Finally, the battery supplier must contribute a deep understanding of energy storage technology, including energy, charge and discharge power capacities and the effect of aging on the battery electrochemistry.

Combined with modelling, these factors determine the cost profile, which is made up of operating revenues and penalties to balance lifetime costs, asset lifetime costs, operating expenses and capital expenditures.

Modelling is an iterative process that starts with a first estimate of battery specification that is combined with the other inputs to the EMS to deliver a cost profile. By repeating the process with a range of sizes, it’s possible to identify a sweet spot, where the operator will find the optimum balance between revenues and costs during the installation’s entire life time.

Understanding Battery Performance

In addition to sizing the ESS correctly, it is equally important to understand the factors that lead to high performance and a long and predictable life.

Good thermal management is the most crucial factor and ensures the temperature is consistent across the entire ESS. By minimizing temperature variation, the cells and modules experience a constant rate of aging. In turn, this allows for precise prediction of battery performance over its lifetime.

Other important aspects are to ensure accurate measurement of state of charge (SOC), good SOC management and high energy efficiency of the battery system itself, as well as the power converter and auxiliary systems such as cooling plant.

Together these extend ESS lifetime, enhance performance and optimize the total cost of ownership.

Optimum ESS for Puerto Rico

A 10 MW PV plant in Puerto Rico successfully optimized an ESS (Figure 7). The grid code required that ramp rates be limited to no more than 10 percent in output change per minute. In addition, the operator required support to grid frequency by up to 5 percent.

FIGURE 7: Schematic of 10 MW PV plant in Puerto Rico
FIGURE 7: Schematic of 10 MW PV plant in Puerto Rico

Modeling was used to identify the optimum ESS as having 1.3 MWh energy storage capacity and 5 MW power rating. Saft delivered a solution comprising three Intensium Max 20 P high power containerized systems. This balanced both the peak power and customer’s requirements. It also anticipated a drop in power over the lifetime of the ESS due to electrochemical aging. This should ensure that the ESS will continue to meet the operator’s minimum technical requirements even at the end of its lifetime.

Endesa Gran Canaria-Ensuring SOC Algorithm Accuracy

A 3 MW ESS has been installed on the Spanish Island of Gran Canaria as part of Endesa’s pioneering STORE (Storage Technologies of Reliable Energy) project to demonstrate how energy storage can promote the integration of renewable energy within utility networks.

A main aim of the STORE project is to demonstrate the technical and economic viability of large-scale energy storage, especially in facilitating the integration of intermittent renewable generation within power networks by making it both predictable and grid compatible.

Saft delivered a fully integrated turn-key ESS based on three Intensium Max 20+ containerized systems. Together, the three units deliver 3 MWh of power to help smooth peak demands on substations and compensate for the intermittent production of wind farms and PV installations, as well as deliver ancillary services to control network frequency and voltage.

The project has confirmed that a SOC algorithm (developed by Saft for its battery management system) corresponds well with the measured energy integration.

The accuracy of the SOC algorithm contributes to:

  • Effective cell balancing
  • Effective use of capacity to generate revenue
  • Permanent operation at partial SOC without need for a maintenance charge

Laboratory tests on a small (56 kWh) battery established a DC roundtrip efficiency averaging 97 percent and better than 95.5 percent for any power rate. This stability at different power rates ensures consistent performance. Field testing on the 3 MWh ESS confirmed the efficiency, showing that performance and functionalities measured on small systems represent full scale operation.

Owners and operators of power plants must have complete confidence that their ESS is consistent, predictable and the right size. To achieve this, they must recognize that ESS operational profiles are complex and multi-functional and sizing is an iterative process involving the battery and EMS.

Modeling is an iterative process that starts with a first estimate of battery specification that is combined with the other inputs to the EMS to deliver a cost profile.


Michael Lippert is marketing and business development manager for Saft’s Energy Storage Business. He holds a degree in European business studies in France and Germany and has been working for more than 20 years in different international sales and marketing positions at Saft focusing on railway, traction and stationary markets. His current responsibilities cover strategic and operational marketing for industrial battery markets, in particular market and product development for renewable energies and smart grids.