ROI Optimisation for Solar PV and Battery System
- Bryan Ellama
- 2 days ago
- 15 min read
This research examines the financial and energy-saving potential of different battery management approaches when paired with photovoltaic (PV) systems in an office building located in Kuala Lumpur, Malaysia. It addresses the common issue of timing inconsistencies between solar energy production and the building's energy usage, which tends to peak during the morning and afternoon hours. To explore this, a comprehensive simulation was conducted using IESVE alongside a Python-based algorithm, allowing for the modeling of five different PV sizes (ranging from 80% to 120% of peak demand) and four distinct management strategies: utilizing PV alone, integrating a standard battery with PV, employing a smart battery with cost-based optimization, and implementing a smart battery with setpoints. The findings reveal that the smart battery approach can lead to an impressive annual cost reduction of up to 45% and offers the quickest return on investment (4.3 years). The study highlights that the key factor influencing the economic viability of PV-storage systems in office environments is the use of intelligent algorithms, rather than merely the size of the system itself. Additionally, an analysis of marginal returns shows that the ideal battery capacity is heavily influenced by the existing storage costs, indicating that larger capacities (40 kWh) only become financially sensible when paired with smart control systems and reduced battery prices (below 1500 RM/kWh).
INTRODUCTION
The integration of renewable energy systems, particularly solar photovoltaics, into commercial buildings plays a vital role in the global shift towards sustainable energy. In office environments, however, there's a notable challenge: solar energy generation peaks around midday, while energy consumption tends to be highest in the morning and late afternoon. This timing mismatch restricts the extent to which photovoltaic systems can utilize the energy they produce. One solution to this problem is battery storage, which can help align energy availability with demand over time. The economic viability of battery systems is affected by their capacity and the sophistication of the energy management system (EMS) that controls them.
Although earlier research has explored the differences between photovoltaic systems with and without battery storage (Hoppmann et al., 2014; Barbosa et al., 2022), there has been less focus on quantifying the benefits of intelligent battery management systems that can adapt to fluctuating electricity prices and anticipated energy needs (Parra et al., 2017). This study seeks to bridge that gap by conducting a comprehensive economic analysis of various battery control strategies, aiming to uncover how smart energy management can enhance profitability in the context of a typical tropical office setting.
MATERIALS AND METHODS
Building Model and Load Profile
In Kuala Lumpur, a typical office building was designed with SketchUp and analyzed using IESVE 2025. This model reflects the daily rhythm of its occupants, operating from 7:00 AM to 6:00 PM, Monday through Friday, while accounting for the essential internal loads such as heating, ventilation, and air conditioning (HVAC), lighting, and IT equipment. After running simulations, it was found that this office consumed approximately 54,083.7 kWh of electricity over the course of the year 2024, highlighting the energy demands of a busy workplace.
PV and Battery Configuration
Five different capacities for photovoltaic (PV) systems were explored: 32.16, 36.18, 40.2, 44.2, and 48.4 kWp, which annually can produce 80% to 120% of the building's annual energy consumption. Each of these configurations was carefully matched with a lithium-ion battery, with capacities ranging from 17.7 kWh to 29.5 kWh, ensuring a proportional relationship. These batteries exhibited an impressive round-trip efficiency of 95%, with charge and discharge rates varying between 4.2 and 8.0 kW, and a depth of discharge also at 95%. The estimated investment costs were set at 3,000 RM per kWp for the PV systems and 1,000 RM per kWh for the batteries.
To determine the ideal battery capacity for the 80% photovoltaic (PV) scenario, we applied a straightforward sizing principle: the battery should be just capable of being fully charged by the surplus energy generated from the PV system under normal conditions. This method promotes efficient use of resources without going too far on capacity. In this case, we found that a 17.7 kWh battery would perfectly align with the maximum daily surplus energy from the PV system, which is about 18 kWh. This setup allows the battery to be fully charged using solar energy while keeping any excess capacity to a minimum. To maintain consistency, we ensured that this proportional relationship was reflected across all PV system sizes, with battery capacities increasing in a linear fashion from this initial baseline.
The sizing of the photovoltaic (PV) system, which in this case corresponds to a full capacity of 40 kWp, is determined by the total energy needs of the building over the course of a year. Although the maximum power demand at any given moment is 18 kW, we opt for a higher peak power for the PV system to accommodate a couple of important factors:
First, we must consider the weather's unpredictability and that sunlight can be suboptimal.
Second, it's crucial to produce enough extra energy during sunny hours to ensure the battery can be charged adequately.
By taking this approach, we can guarantee that the system will fulfill the building's energy requirements throughout the year, rather than just meeting the peak demand at a single moment.
Tariff Structure and Economic Parameters
The cost of electricity was approached as a combination of an energy charge (0.32816 RM/kWh) and a demand charge (59.84 RM/kW based on the highest monthly usage). To determine the return on investment (ROI), we defined it as
where Annual Savings = (Base case cost – Scenario cost).
Energy Management Scenarios
Five scenarios were simulated:
Case 1: Reference (no PV)
Case 2: PV only
Case 3: PV + standard battery (charge from PV surplus only)
Case 4: PV + smart battery (cost-based optimization using grid and PV)
Case 5: PV + smart battery with charge setpoints (predefined SOC thresholds before peaks)
An algorithm developed in Python mimicked the hourly operations over a year, tracking solar energy production, the battery's charge levels, electricity drawn from the grid, and the overall energy expenses.
The estimated cost of the battery is around 1000 RM per kWh, and we present a sensitivity analysis of this figure in Section 3.5.
RESULTS
Table 1 summarizes the main results for all PV capacities and management cases.
Performance by Management Strategy
Across all PV capacities, the smart battery (Case 4) stood out by delivering the greatest annual savings while also having the lowest return on investment (ROI). When compared to the standard battery (Case 3), the smart algorithm was able to cut annual energy expenses by as much as 45%. This impressive reduction was largely due to its clever peak shaving capabilities, which effectively lowered demand charges. On the other hand, the setpoint-based control (Case 5) did enhance power smoothing, but it occasionally resulted in less-than-ideal economic outcomes.
Performance by PV Capacity
As we increased the capacity of photovoltaic (PV) systems, we noticed that the benefits began to diminish once we surpassed 100% of peak demand. The sweet spot for balancing our investments and the savings we could achieve was found in the range of 80% to 100% of the PV size when paired with a smart battery. This combination allowed us to enjoy a return on investment (ROI) that typically fell between 4.3 and 4.5 years.
Battery economic analysis
To assess the economic feasibility of various battery capacities, we undertook a complete marginal return analysis. In Figure 8, we present the marginal returns associated with upgrading from the standard 17 kWh battery to a more robust 40 kWh option following the “Standard 1 kWh per 1 kWp rule”, considering different sizes of photovoltaic (PV) systems and battery costs. The findings highlight distinct economic thresholds: when battery costs are lower (between 1000-1250 RM/kWh), the marginal returns range from 12-24%, making the larger capacity an appealing choice, especially for smart battery systems (Cases 4-5). However, as costs rise, these returns significantly decrease, dropping below 8% at a price of 1750 RM/kWh.
Table 2 further details this economic relationship and offers specific recommendations based on battery costs. The 17 kWh system represents the capital-efficient solution, while the 40 kWh system becomes viable only under favorable cost conditions and when paired with intelligent control strategies.
Table 2: Battery capacity economic recommendations based on marginal return analysis
Battery cost (RM/kWh) | Average Marginal Return - Case 4 (%) | Economic recommendation |
1000 | 15-24 | 40 kWh favorable |
1250 | 12-18 | 40 kWh attractive |
1500 | 8-12 | Borderline |
1750 | 4-8 | 17 kWh recommended |
2000 | 0-4 | 17 kWh strongly preferred |
Cost sensitivity and peak demand analysis
The ROI sensitivity analysis (Figure 9) illustrates the significant influence of battery costs on investment returns for different capacity scenarios. The 17 kWh system stands out with an impressive ROI of 4.3–4.9 years across most cost scenarios, while the 40 kWh system only matches these returns at the lowest price level of 1000 RM/kWh. This highlights the capital efficiency of the smaller system in today's market landscape.
When we look at peak demand reduction performance (Figure 10), it becomes clear why smart battery strategies are economically advantageous. Case 4, featuring the smart battery, achieves a remarkable 65-75% reduction in peak demand across all PV sizes, far surpassing the standard battery (Case 3), which only manages a 45-55% reduction. This impressive peak shaving capability translates directly into savings on demand charges, accounting for 60-70% of the total economic benefits.
The cost sensitivity analysis (Figure 11) reinforces the strength of smart battery strategies at various price points. Although absolute ROI values tend to rise with increasing battery costs, the smart battery (Case 4) consistently enjoys a 0.8-1.2 year edge over other strategies, showcasing the adaptability of intelligent control in response to market changes.
DISCUSSION
Interpretation of Results
The findings highlight that battery storage in tertiary buildings offers more than just a way to keep excess PV energy. With smart control algorithms, we can effectively manage demand charges and cut costs by charging from the grid when electricity is cheaper. This proactive approach not only boosts returns on investment but also provides greater flexibility in operations.
Comparative Analysis of Smart Battery Strategies: Case 4 vs Case 5
The nuanced performance difference between Case 4 (smart battery with cost optimization) and Case 5 (smart battery with setpoints) reveals important insights into optimization trade-offs.
Case 4's superior economic performance arises from its purely cost-driven algorithm that dynamically balances energy charges against demand charges without predefined constraints. By charging from the grid exclusively during the lowest-cost periods and discharging during peak demand hours, it achieves optimal economic arbitrage. Artificial State of Charge (SOC) thresholds cannot constrain the system's ability to minimize total costs due to its flexibility.
Case 5's setpoint strategy, while effective for peak shaving and load profile smoothing, introduces operational constraints that can limit economic optimization. The predefined SOC thresholds (e.g., maintaining 70% charge before expected peaks) force premature grid charging, sometimes during suboptimal tariff periods. This "safety margin" approach ensures reliable peak shaving capability but comes at the expense of increased energy costs, particularly when:
Grid charging occurs during intermediate tariff periods rather than strictly during lowest-cost windows
Excess battery capacity is maintained that could otherwise be utilized for additional economic arbitrage
Conservative setpoints lead to underutilization of the battery's full economic potential
The differences in performance become particularly noticeable when solar panel capacities are at medium levels (80-100%), where energy distribution is more even. As we move to higher solar energy usage (120%), the gap starts to close because the excess solar energy lessens the reliance on grid electricity, making the limitations of the setpoint strategy less financially taxing.
This analysis indicates that while using setpoint-based control can enhance grid stability and help manage energy loads predictably, algorithms focused solely on cost optimization tend to yield better financial outcomes, especially in situations where demand charges and fluctuating energy prices play a significant role.
Comparison with Literature
Recent research, including the work of Parra et al. (2017), Hoppmann et al. (2014) and Barbosa et al. (2022), underscores the idea that smart battery management can significantly boost profits when using time-of-use tariffs. In regions like Malaysia, where the sun shines brightly and the energy pricing system has significant demand charges, adopting such smart strategies can lead to substantial financial benefits.
Economic Optimization: PV Sizing Sweet Spot and Battery Capacity Trade-offs
Identifying the perfect range for sizing photovoltaic (PV) systems, particularly between 80% and 100% of peak demand, provides helpful guidelines for practical use. This ideal range, commonly known as the "sweet spot," emerges from a complete economic and technical factors that result in diminishing returns when the capacity surpasses 100%.
Key Factors Influencing the Sweet Spot:
Finding the right balance in sizing photovoltaic (PV) systems, especially within the 80% to 100% range of peak demand, is essential for practical application. This sweet spot is where economic and technical factors converge, leading to diminishing returns when capacity exceeds 100%.
When we discuss load-profile alignment, operating within this capacity range means the system can effectively mirror the building's energy needs during those sunny peak hours. The smart battery is like a helpful assistant, fine-tuning any discrepancies during the morning and evening rush, which boosts self-consumption and minimizes energy waste.
Battery efficiency is another key aspect. In this optimal range, the battery operates at its peak cycling rate, avoiding the risks of underutilization that smaller PV systems might face, as well as the excess energy overload that larger systems can produce. By aligning battery capacity with the PV system size, we ensure that energy flows are both balanced and efficient.
In addition, to effectively reduce demand charges, keeping the capacity between 80% and 100% ensures that there is sufficient energy to charge the battery for peak shaving. This approach helps in managing costs and prevents the financial drawbacks associated with excessive grid exports that fail to lower demand charges.
Understanding Economic Optimization:
The idea of diminishing returns becomes clear when we examine PV capacity that goes beyond 100%, and we can explain this through a closer look at marginal analysis:
For the smart battery case (Case 4):
80% → 90% PV: Marginal return = 21.3%
90% → 100% PV: Marginal return = 15.7%
100% → 110% PV: Marginal return = 16.6%
110% → 120% PV: Marginal return = 11.2%
The marginal return of 11.2% when boosting capacity from 110% to 120% signifies a crucial turning point. Beyond this threshold, further investments typically yield less attractive returns, especially when compared to the more appealing range of 15.7–21.3% seen at lower capacity levels.
Battery Capacity Economic Boundaries:
The analysis of marginal returns reveals that today's storage costs have a significant impact on choosing the optimal battery size. Although the 17.7 kWh system stands out as the most cost-effective option for our specific case, the financial feasibility of larger battery capacities is defined by distinct thresholds:
- At lower battery costs (1000-1250 RM/kWh), marginal returns of 15-24% make the 40 kWh capacity economically attractive, particularly for smart battery systems (Cases 4-5)
- At moderate costs (1500 RM/kWh), returns of 8-12% make the decision borderline and dependent on specific investment criteria
- At higher costs (1750+ RM/kWh), the 17.7 kWh system clearly represents the optimal capital-efficient solution
The findings reveal that intelligent control acts as a powerful economic enhancer. For basic batteries (Case 3), the additional benefits from expanding capacity are quite limited, ranging from 0 to 7%. In contrast, smart batteries (Cases 4-5) can tap into significant value, yielding returns of 15-24% when larger capacities are utilized at reasonable costs.
Practical Considerations for System Design:
Risk Management: Keeping the PV capacity between 80 and 100% is a smart strategy to avoid overestimating building loads or relying too heavily on anticipated future load reductions. By pairing this with the 17.7 kWh battery, even at the current higher costs, we can ensure that the system remains financially sound, regardless of changing circumstances.
Optimizing Grid Interaction: Systems that function within this ideal capacity range help lighten pressure on the grid. They effectively balance the need to minimize energy imports while managing exports, which is in line with utility companies' growing preference for decentralized energy solutions.
Future-Proofing Strategy: The capacity analysis provides a clear roadmap for system expansion—as battery costs decline below 1250 RM/kWh, upgrading to 40 kWh capacity becomes economically justified while maintaining the same smart control.
This optimization approach goes beyond simple recommendations for the 'optimal size.' It equips designers with flexible decision-making tools that weigh capital efficiency against the potential for maximum savings, with specific market conditions and financial limitations.
Analysis of the PV Sizing Sweet Spot: Economic Optimization Boundaries
Finding the ideal range for photovoltaic (PV) system sizing, specifically between 80% and 100% of peak demand, is a significant insight for real-world applications. This optimal zone, often referred to as the "sweet spot," arises from a combination of various economic and technical elements that lead to diminishing returns when exceeding 100% capacity.
Key Factors Influencing the Sweet Spot:
Load-Profile Alignment: Operating at 80-100% PV capacity allows the system to closely match the building's energy usage during peak sunlight hours. The smart battery plays a crucial role by adjusting for any remaining differences during morning and evening peaks, thereby enhancing self-consumption without major energy waste.
Battery Efficiency: In this capacity range, the battery functions at its best cycling rate—avoiding both underuse, which can occur with smaller PV systems, and the overload of excess energy typical of larger systems. By proportionally scaling battery capacity to the size of the PV system, energy flows remain balanced and efficient.
Minimizing Demand Charges: The 80-100% capacity range ensures there is enough energy to charge the battery for peak shaving, while also steering clear of the financial drawbacks associated with excessive grid exports that do not aid in reducing demand charges.
Economic Optimization Mechanism:
The concept of diminishing returns becomes evident when we look at PV capacity exceeding 100%, and we can explore this through marginal analysis:
For the smart battery case (Case 4):
80% → 90% PV: Marginal return = 21.3%
90% → 100% PV: Marginal return = 15.7%
100% → 110% PV: Marginal return = 16.6%
110% → 120% PV: Marginal return = 11.2%
The marginal return of 11.2% from increasing capacity from 110% to 120% marks an important juncture. After reaching this point, additional investments often result in less appealing returns compared to the more enticing range of 15.7–21.3% observed at lower capacity levels.
Practical Implications for System Design:
Risk Mitigation: Staying within the 80-100% capacity range serves as a protective measure against overestimating building loads or expecting future load reductions. This approach helps ensure that the system remains financially viable, even as circumstances change.
Grid Interaction Optimization: Systems operating within this ideal range contribute to reducing strain on the grid. They strike a balance between minimizing energy imports and effectively managing exports, which aligns with utility companies' preferences for decentralized energy generation.
Space and Resource Efficiency: Thoughtful sizing of systems avoids unnecessary expansion of roof-mounted installations, preserving space for other uses and reducing material consumption.
This analysis of the optimal capacity range offers valuable insights for designers and investors alike. It reinforces the idea that "bigger is not always better" in the design of PV-battery systems, emphasizing that smart control strategies can unlock the full potential of systems that are appropriately sized.
Limitations and Future Work
This study assumes fixed the costs associated with investments and straightforward pricing models. In reality, fluctuations in electricity pricing frameworks, battery degradation, and weather uncertainties could alter profitability. Future work should include:
- Sensitivity analysis on tariff and investment parameters;
- Integration of weather and load forecasting into the EMS;
- Lifecycle modeling including battery degradation;
- Multi-building optimization to assess aggregated demand management potential.
CONCLUSION
This study highlights how smart battery management can significantly enhance the economic feasibility of photovoltaic (PV) systems in commercial buildings. The clever algorithms integrated in energy management systems (EMS)—especially those focused on optimizing demand charges—offer more value than merely expanding the system's size. The smart battery approach (Case 4) achieved a remarkable cost reduction of up to 45%, with a return on investment (ROI) of less than 4.5 years for systems that are optimally sized (80-100% of demand).
Importantly, the economic analysis shows that designing an optimal system requires a multi-faceted approach. While intelligent control mechanisms consistently outperform traditional methods, the ideal battery capacity is influenced by current storage costs and investment strategies. The 17.7 kWh system stands out as the most cost-effective solution, especially given the current high prices of batteries, while the 40 kWh system provides a forward-looking option as storage costs are expected to decrease.
For those in decision-making roles, this research offers a practical framework suits tropical office buildings. It underscores that investing in sophisticated EMS algorithms is a powerful step toward achieving cost-effective decarbonization. System designers are encouraged to use marginal return analysis based on real-time storage costs to identify the most economically viable configuration for their unique situations, ensuring both environmental responsibility and financial health.
Acknowledgement
This work was done in collaboration with Simon Laporte, who especially provided valuable guidance on the python programming and optimisation, as well as by Gregers Reimann on the conceptual framework of the study.
References
Hoppmann, J., Volland, J., Schmidt, T. S., & Hoffmann, V. H. (2014). The economic viability of battery storage for residential solar photovoltaic systems—A review and a simulation model. Renewable and Sustainable Energy Reviews, 39, 1101–1118.
Barbosa, L. S., Mendes, A. P., & Dantas, G. A. (2022). Impact of battery storage on residential PV self-consumption: A review. Journal of Energy Storage, 49, 104095.
Parra, D., Norman, S. A., Walker, G. S., & Gillott, M. (2017). An updated review of the optimal planning of photovoltaic-battery systems. Progress in Energy and Combustion Science, 56, 1–19.
Lazard. (2023). Levelized Cost of Storage Analysis—Version 9.0. Lazard Ltd.