Unlocking Precision: Powin’s SOC Algorithm Redefines Energy Estimation Accuracy

By Kyle Smith, Sales Director, Monique Wong, Sr. Product Manager, Eric Stone, Powin, Sr. Product Manager

Read the full 2024 Solar Risk Assessment Report at: https://www.kwhanalytics.com/solar-risk-assessment


State of Charge (SOC) represents a Battery Energy Storage System’s (BESS) available energy for discharge. SOC is critical in predictably committing to dispatch schedules and can lead to penalties if commitments for delivery of grid services cannot be fulfilled due to insufficient energy/capacity. Unfortunately, poor SOC estimation is common. Lithium Iron Phosphate (LFP) cells exhibit a flat voltage-SOC curve (Figure 1), making accurate estimations based on voltage alone nearly impossible between 20-95% SOC. The specific electrical architecture connecting thousands of cells further complicates the problem and reduces SOC certainty at the site level.

The conventional method of calculating SOC is referred to as “Coulomb Counting”. This method is prone to accumulating error since it relies on accurately measuring charge / discharge current over time and an accurate initial SOC reading. Conventional SOC algorithms do not correct for sensor error, nor do they account for the electrical architecture connecting cells into a BESS. Powin has measured up to an average of 7% SOC error in field testing using conventional methods.

State of Charge
Figure 1: Voltage vs. SOC Relationship for LFP cells (an example)

Powin addresses these issues with an advanced SOC algorithm that leverages terabytes of cell-level data from Powin’s operational fleet. The advanced SOC algorithm combines Coloumb Counting with data-driven state estimations to adjust for sensor error, temperature, and the effects of current on voltage. At the string level (individual unit/rack with capacity of around 230-375 kWh), this algorithm achieves an average error of only 3%, based on more than 6 months of operational studies (Figure 2).

Figure2
Figure 2: SOC Error - Conventional Method vs Powin’s Algorithm

In summary, accurate State of Charge (SOC) estimates are essential for the reliable operation of BESS assets and their ability to provide grid services effectively, Powin’s advanced SCO algorithm increases the accuracy of estimated energy available for discharge, a necessary update to current technology as the industry works to improve the energy and power available of all BESS systems with the goal of increased operation and financial reliability.

For the latest updates and news please visit our website or connect with us on LinkedIn.