Case Study
Battery analytics in African C&I solar projects
1
Introduction
Reliable electricity supply across emerging markets is increasingly dependent on the adoption of battery storage systems. Batteries represent a large part of overall system costs. Early battery failures can hence be a big pain point for system operators and financiers. The main reasons for reduced service life are high battery temperatures and poor operation strategies.
Poor network quality is all too common across emerging markets and can lead to poor data quality. Therefore, one goal of this case study was to see how health-monitoring algorithms could deal with imperfect data sets.
The energy service provider Pirano Energy, the battery analytics provider ACCURE and remote monitoring provider AMMP Technologies collaborated to identify the potential of advanced battery analytics for extending the service life of batteries in real-life applications. The impact analysis was done using operational data of 14 PV-battery systems for C&I applications in across Nigeria.
Opportunities for battery analytics
For C&I solar projects, batteries often make up around
of the total replacement costs
0
%
2
Participants
3
Case study
ACCURE, AMMP, and Pirano collaborated to gain field experience on how advanced battery analytics can support Energy service companies in the C&I market in operating and managing their batteries.
Goals
- Determine feasibility of battery analytics for African C&I solar projects
- Identify value of advanced battery analytics for energy service companies like Pirano
Scope
This case study has used historical battery data from 14 small scale grid-tied C&I projects in Nigeria.
C&I systems in Nigeria
0
PV capacity
3 to
0
kW
Lithium-ion batteries
4.8 to
0
kWh
Data quality
The historic battery data is acquired and standardized by AMMP over a period of 11 months and contains irregular data outages due to power or connection outages.
The data includes the pack voltage, pack current and battery SOC in a resolution of 5 mins, including irregular data outages.
| Parameters | Acual data |
|---|---|
|
Data resolution |
5 min
|
|
Period of data |
11 months (including data outages)
|
|
Physical quantities |
– Pack voltage – Pack current – SOC – Temperature |
Feasibility of battery analytics for African C&I solar projects
For many African solar projects, an unstable internet connection is still a major problem and can lead to poor data quality. To assess whether state-of-the-art battery analytics algorithms can still perform as desired, low data quality is chosen in this case study, as well.
The aging and operational analysis below examine the insights that battery analytics can generate based the historical battery data of Pirano's systems.
The analysis can be grouped into:
Aging analysis
Operational analysis
Aging analysis
By evaluating the historical battery voltage and current, ACCURE can derive the remaining capacities of Pirano's battery systems and thus determine the State of Health (SOH) at each point in time.
Note
For the best experience with the embedded interactive charts, you can use Google Chrome or temporarily turn off "Prevent cross-site tracking" on Safari.
This interactive chart shows the aging analysis results of seven exemplary battery systems stationed in Lagos, Nigeria. While most observed batteries expectedly age by a small single-digit percentage amount per year, system 2 stands out by having lost 10% of its initial capacity within just a year.
Assuming there was no change in operating conditions, it would imply that the battery used in system 2 could reach its end-of-life in as little as two years in total.
Further, the analysis allows early and clear identification of warranty cases. If the SoH drops below 80 percent while the battery is still under warranty, a replacement should be demanded from the supplier. ACCURE’s analysis proves that the battery was always operated within its specifications. For System 2 this means that it will be eligible for a warranty replacement soon.
The state of health (SoH) helps to:
Identify batteries that are most likely to fail soon
Identify warranty cases
Operational analysis
Historical data is a great source of information that can be used to extend the lifetime of a battery by optimizing the operation strategy. The evaluation focuses on state of charge (SoC), Voltage and Temperature characteristics.
As the battery of system 2 is exposed to the most accelerated aging, the operation of this battery is analysed in detail.
The evalutaton in this section focuses on:
- State of Charge (SoC)
- Temperature
State of charge (SoC)
The SoC histogram of system 2 shows that the system is fully charged most of the time (1650 hours), while seldom discharged.
A forecast-based charging strategy could reduce the stress on the battery and thus prolong its lifetime. Load forecasts in combination with battery aging predication can be used to set dynamic charging limits without reducing the utility of the battery.
Battery temperature
Another cause for increased capacity loss can be a high battery temperature. In terms of aging, temperatures above 20°C should be avoided. The diagram in Figure 9 shows the temperature distribution of System 2.
It’s clear that the temperature of the battery is excessive throughout the entire sample period, presenting a second potential root cause of accelerated battery aging for System 2. Improved cooling could help to improve the battery lifetime.
Closing remarks
Despite the limited data granularity, AMMP’s monitoring + ACCURE’s battery intelligence was able to:
1. Identify batteries which are exposed to accelerated aging
2. Understand root causes of accelerated aging
3. Suggest changes of operation strategies
The value of battery analytics: Pirano's perspective
Stephen Bello
Pirano Energy
Jonathan Schulte
AMMP Technologies
Marc Mennekes
ACCURE Battery Intelligence
After delivering the results of the battery analysis, AMMP’s Jonathan Schulte and ACCURE’s Marc Mennekes interviewed Stephen Bello, Pirano’s Technical Customer Support Lead
Jonathan: Which of the generated battery insights are stood to you?
Stephen: The insights drew our attention towards batteries with degrading State of Health in comparison to the number of charge/discharge cycles. The high operating temperature was identified as a root cause hence the need to improve the cooling around the batteries to improve their lifespan.
Marc: How can you explain the high operating temperature of some of your batteries?
Stephen: For the system with the highest operating temperature and interestingly also the lowest SOH (system 2), we increased the battery inverter capacity from 5 kW to 10 kW last year. The additional excess heat led to an ambient temperature jump from 38 °C to 54 °C in the operation room.
Note: A high ambient temperature can lead to an increase in battery temperature.
Jonathan: Can you take any measurements to prevent the battery from being exposed to these high temperatures?
Stephen: Our cooling solutions are very customized; they depend on the surroundings and the client’s desires. For this specific system, it is not possible to add any additional cooling solution. We can only ensure that the windows and doors of the operation room remain open as much as possible.
Nevertheless, after seeing the clear correlation between the high battery temperature and battery aging, we will design our systems and include cooling solutions such as fans or air conditions more frequently.
Marc: Would you recommend the use of advanced battery analytics?
Stephen: Yes. The granular overview of battery insights helps our team to manage and operate the batteries more efficiently while extending the lifetime of batteries. This finally reduces the cost of maintenance and expenditure on changing the batteries.
4
Conclusion
The focus of this work was to illustrate the feasibility and benefits of a battery analytics solution for PV-battery systems in C&I projects in emerging markets.
As is common for PV-battery projects in emerging markets, the data in this case study had limited granularity and several data outages. Despite the impaired dataset, ACCURE’s algorithms were able to perform all insightful analytics. Through computing the State of Health (SOH) and analyzing the historic operation, the battery analytics enabled us to achieve:
| Result | Example in the case study |
|---|---|
|
Identified batteries which are exposed to accelerated aging |
Battery of System 2 losing 10% capacity per year. In an ideal scenario, it should be 2-4% |
|
Understand the root causes of accelerated aging |
Battery of System 2 exposed to high temperatures |
|
Suggests changes of operation strategies |
Change cooling strategy for system 2 |
High temperatures and high SOCs led to higher aging rates (capacity loss per year). For the system with the highest capacity loss, the main root cause was identified as high operating temperatures.
After sharing these results, Pirano was able to confirm that this single system is operating in bad conditions. The excess heat of the electrical equipment was responsible for heating up the operation room and consequently the battery.
With these insights, Pirano was able to change the cooling strategy of the battery and can use these findings for their future storage projects.
A successful collaboration
This study led to some really valuable insights for Pirano’s operations. AMMP can confidently endorse ACCURE’s battery analytics services and offers them as an add-on to its platform.
To ensure the correct operation and performance of the battery systems in a C&I solar projects, it is important to monitor and analyze the batteries during the service life. If enhanced battery analytics are used for supervising a whole battery portfolio, it is easier to process the maintenance or replacement battery systems when a critical aging state (knee-point) is reached.
A continuous monitoring of the whole portfolio can help save on replacement, maintenance, and warranty costs.
