LCOS vs CAPEX:
How much does energy storage really cost?
"How much is a battery?" -- this is the wrong question. The right question for an investor is: "How much will 1 kWh cost me, which I will get from this battery in 5 years?". The answer to this is provided by the LCOS (Levelized Cost of Storage) indicator -- the normalized cost of energy storage, which takes into account the entire life cycle of the system.
LCOS is an analogue of LCOE (Levelized Cost of Energy) for generating capacities, adapted for storage systems. The methodology is actively used by BloombergNEF, Lazard, IRENA and other analytical institutions to compare different energy storage technologies on an equal basis.
The low CAPEX trap
Many companies choose BESS based on the principle of "lowest purchase price" (CAPEX). They buy lead-acid (AGM/GEL) or cheap lithium batteries from unknown manufacturers. This is a fatal mistake that costs Ukrainian business millions of hryvnias every year.
Why? Because in the energy sector we do not buy a "box", but number of cycles. A cheap battery with a resource of 500 cycles is more expensive than an expensive battery with 6,000 cycles, if you consider the cost of energy given. It's counterintuitive, but math is merciless.
Lead-acid batteries have three critical limitations that dramatically increase their LCOS: limited depth of discharge (50% vs. 90% in LiFePO4), lower efficiency (80-85% vs. 95-97%), and exponential decline in life as depth of discharge increases. If you discharge the AGM by 80%, it will last not 500, but 200-250 cycles.
Expanded LCOS formula
Simplified LCOS calculators can be misleading. The full formula takes into account the time value of money (discounting), capacity degradation, module replacement costs and the cost of charging energy.
// FULL LCOS FORMULA
LCOS = (CAPEX + NPV(O&M) + NPV(Charging) + NPV(Replacement)) / NPV(Total Energy Discharged)
CAPEX -- initial investments (equipment + installation + design + PCS + BMS)
NPV(O&M) -- the reduced cost of operating expenses for the entire period:
SUM[ O&M(t) / (1 + r)^t ] for t = 1..N
NPV(Charging) -- given cost of electricity for charging:
SUM[ E_charge(t) * Price_elec(t) / (1 + r)^t ]
NPV(Replacement) -- the cost of replacing modules or augmentation is given:
SUM[ Repl_cost(t) / (1 + r)^t ] (if capacity < replacement threshold)
NPV(Discharged) -- the given volume of the given energy of taking into account the degradation:
SUM[ Capacity(t) * DoD * Efficiency * Cycles(t) / (1 + r)^t ]
r -- discount rate (typically 8-12% for Ukraine)
N -- evaluation horizon (10-20 years)
Capacity(t) -- nominal capacity of taking into account degradation: C0 * (1 - deg_rate)^t
The key point: the discount rate for Ukrainian projects is significantly higher than the European one (8-12% versus 4-6%). This means that technologies with a longer service life receive an even greater advantage in LCOS analysis, because future cash flows are discounted more strongly, and the initial investment is spread over a larger amount of delivered energy.
Comparison of 4 technologies of accumulation
Expanded comparison for a system with a nominal capacity of 100 kWh. All LCOS calculations are carried out taking into account degradation, inverter efficiency and a discount rate of 10%.
| Parameter | AGM | GEL | NMC | LiFePO4 |
|---|---|---|---|---|
| CAPEX (100 kWh) | $12,000 | $16,000 | $25,000 | $28,000 |
| Resource (cycles at DoD recom.) | 400-600 | 800-1200 | 3000-4000 | 6000-8000 |
| Recommended by DoD | 50% | 50% | 80% | 90% |
| Cycle efficiency (round-trip) | 78-82% | 80-85% | 92-95% | 95-97% |
| Degradation per year | 8-15% | 5-10% | 2-3% | 1-2% |
| Operating temperature range | -20...+50 C | -20...+50 C | 0...+45 C | -10...+55 C |
| Calendar resource | 3-5 years | 5-8 years | 8-12 years | 15-20 years |
| LCOS ($/kWh) | $0.72 | $0.38 | $0.12 | $0.06 |
LCOS technology comparison ($/kWh per cycle)
Na-Ion (sodium-ion) batteries -- forecast LCOS based on CATL and HiNa Battery data for series production of 2027 years.
Factors affecting LCOS
C-Rate and its effect on degradation
C-Rate (the ratio of the charge/discharge current to the nominal capacity) is one of the key factors that determines the rate of degradation. For LiFePO4 at 0.5C, the resource is 6000+ cycles, but with constant operation at 2C, this indicator can decrease to 3000-4000 cycles. For NMC, the situation is even worse: at 2C, the resource drops by 40-50% compared to the nominal one.
In practice, this means that for high C-Rate applications (for example, FCR in the market of auxiliary services) it is necessary to either choose an oversized system or include increased degradation in the LCOS calculation.
Temperature influence on the resource
Every 10 degrees above the optimal temperature (25 C) reduces the calendar resource of lithium batteries by approximately 50% according to the Arrhenius rule. In Ukrainian conditions without an HVAC system (container heating/conditioning), the temperature of the batteries can reach 45-50 C in the summer, and drop to -15 C in the winter.
According to research by Sandia National Laboratories, LiFePO4 loses an additional 3-4% of capacity per year when continuously operated at 40 C instead of 25 C. For a system with a horizon of 15 years, this is the difference between 80% of residual capacity and 55%. Therefore, the investment in thermal control ($15,000-25,000 for a containerized solution) pays off in reduced LCOS.
Depth of discharge (DoD) and cycle life
The relationship between the depth of discharge and the number of cycles is non-linear. Here are typical values for LiFePO4 (EVE LF280K, CATL):
Resource of LiFePO4 cycles depending on DoD
Data is approximate, based on EVE Energy and CATL specifications for LFP 280-314 Ah cells. Actual life depends on temperature and C-Rate.
Paradox: at 90% DoD we get fewer cycles (3,500), but more useful energy of each cycle. The optimal DoD of LCOS for LiFePO4 is usually in the range of 70-85%. At the same time, for lead-acid batteries, the optimum is 30-50%, because their degradation curve is much steeper.
Inverter efficiency (PCS)
The value of round-trip efficiency includes not only the chemical efficiency of cells, but also losses in the Power Conversion System (PCS). A typical Deye/Sungrow/Huawei inverter has an efficiency of 97-98% in each direction, which gives an additional 3-6% of losses to the overall round-trip efficiency of the system. With daily cycling, this is equivalent to 3-6% excess power consumption, which directly increases LCOS.
LCOS in the context of Ukrainian tariffs
To assess the feasibility of investments in BESS, it is necessary to compare the LCOS of the tariff spread (the difference between peak and night tariffs). If LCOS is below the spread -- Peak Shaving becomes profitable.
// BREAK-EVEN CALCULATION
Peak tariff (class 2, >150 kW): ~5.20 hryvnias/kW*h
Night rate (coefficient 0.4): ~2.08 UAH/kW*h
Tariff spread: 5.20 - 2.08 = 3.12 UAH/kWh (~$0.075)
Spread of taking into account efficiency (95%): 3.12 - (2.08 * 0.05) = 3.02 UAH/kWh (~$0.073)
LiFePO4 LCOS ($0.06) < Tariff spread ($0.073)
= Peak Shaving PROFITABLE, margin ~$0.013/kWh
AGM LCOS ($0.72) >> Tariff spread ($0.073)
= Peak Shaving is LOSSFUL, loss $0.647/kWh
With a consumption of 500 kWh per day and daily cycling, the net profit from Peak Shaving of LiFePO4 is approximately $0.013 * 500 * 365 = $2,372 per year. With a CAPEX of $28,000, this gives a simple payback period of about 12 years only on arbitration. However, if you add the avoidance of fines for excess capacity and participation in peak programs, the real payback is reduced to 5-7 years.
LCOS vs LCOE (Levelized Cost of Energy)
These two indicators are often confused. The main difference:
LCOE
Standard cost production electricity. It is used for generating capacities (SPP, WPP, TPP). Formula: (CAPEX + NPV(O&M) + NPV(Fuel)) / NPV(Generated Energy).
LCOS
Standard cost storage energy It additionally takes into account charging costs and capacity degradation. BESS does not generate - it stores and gives.
The LCOE of a solar panel in Ukraine is $0.04-0.06/kW*h. LCOS of the storage system adds $0.06/kWh (LiFePO4). Thus, "solar + storage" has a total LCOE of about $0.10-0.12/kW*h -- this is already competitive with the network's peak tariff ($0.13/kW*h). In 2027-2028, due to further lowering of the price of LFP ($80/kWh according to BloombergNEF forecasts), parity will be reached for the basic tariff as well.
Why "cheap" = expensive: A real case
In 2023, the Kharkiv meat processing plant (consumption 800 kWh/day, peak load 200 kW) decided to install a backup power supply system and Peak Shaving. Two options were considered: AGM from a local supplier and LiFePO4 (CATL 280Ah + Deye inverter).
// CASE: Meat processing plant, Kharkiv
Choice A: AGM "economy"
Initial CAPEX: $18,000
Capacity: 150 kWh (DoD 50% = 75 kWh useful)
Resource: 450 cycles (at 50% DoD)
Term of service: ~1.5 years with daily use
First replacement (2024): $16,000
Second replacement (2025): $17,500 (price increase)
Total expenses for 5 years: $51,500
Delivered energy: ~101,250 kWh
LCOS = $0.51/kWh
Choice B: LiFePO4 Tier-1
Initial CAPEX: $38,000
Capacity: 150 kWh (DoD 90% = 135 kWh useful)
Resource: 6,000+ cycles (at 90% DoD)
Term of service: 15+ years with daily use
O&M for 5 years: $2,000 (diagnostics, firmware)
Residual capacity (5 years): ~92%
Total expenses for 5 years: $40,000
Delivered energy: ~233.775 kWh
LCOS = $0.17/kWh
Real case, numbers are rounded. The plant ended up switching to LiFePO4 in 2025 after the second AGM change, losing $33,500 in "savings".
Timeline: When LiFePO4 becomes cheaper than AGM
Year 0 -- Start
AGM: $18,000 spent. LiFePO4: $38,000 spent. AGM seems "more profitable" by $20,000.
Year 1.5 -- First AGM replacement
AGM needs to be replaced. Total AGM expenses: $34,000. LiFePO4 works without interventions, the remaining capacity is 97%. The difference comes down to $4,000.
Year 2.5 -- The Crossing Point
The total TCO (Total Cost of Ownership) of AGM exceeds the TCO of LiFePO4. From that point on, each year of savings is ~$8,000-12,000.
Year 3 -- Second AGM replacement
AGM: $51,500 spent. LiFePO4: $40,000. Difference: $11,500 in favor of LiFePO4. And this is just the beginning.
Year 10 -- Long-term perspective
AGM: 6-7 replacements, $120,000+ spent. LiFePO4: runs at ~85% capacity, costs $42,000 (including O&M). The difference: $78,000. This is the cost of three more LiFePO4 systems.
Comparison: "Cheap" Lead vs "Expensive" LiFePO4
Let's calculate the net economy for a 100 kWh system (basic example without augmentation and discounting).
| Parameter | AGM (Lead) | Tier-1 LiFePO4 (BESS) |
|---|---|---|
| CAPEX (Price) | $15,000 | $30,000 |
| Resource (Cycles) | ~400 | ~6000 |
| DOD (Depth) | 50% | 90% |
| LCOS ($/kWh) | $0.75 | $0.06 |
FAQ: Frequently asked questions about LCOS
Does LCOS factor in the cost of the inverter and BMS?
Yes, if the calculation is done correctly. The full LCOS formula includes all capital costs including Power Conversion System (Inverter/PCS), Battery Management System (BMS), Cabling, Installation, Design and Commissioning. Some manufacturers publish LCOS only for the battery pack, which underestimates the figure by 20-30%. Always ask for "system-level LCOS".
How does inflation affect LCOS?
Inflation affects through two channels. The first is the increase in the cost of O&M and electricity for charging (increases the numerator of the formula). The second - the discount rate usually includes an inflationary component (nominal rate = real + inflation). In Ukraine, with inflation of 8-12% and a real discount rate of 4-6%, the nominal rate reaches 12-18%. This means that technologies with a shorter payback period have an advantage. However, LiFePO4 wins even at a high nominal rate due to its radically longer life.
Is it possible to compare LCOS from different manufacturers?
Only if the same methodology is used. Standardized requirements for calculating LCOS do not yet exist. Main differences: inclusion/exclusion of charging energy cost, discount rate, valuation horizon, degradation assumptions. We recommend either calculating LCOS independently using a single formula, or using data from independent sources (Lazard LCOS Annual Report, IRENA Electricity Storage Outlook).
What is "augmentation" and how does it affect LCOS?
Augmentation is the addition of new battery modules to an existing system to compensate for capacity degradation. For example, if after 8 years the capacity has decreased from 100 kWh to 80 kWh, you can add 20 kWh of new modules. This increases CAPEX (and therefore LCOS), but allows the system to continue operating without a complete replacement. For LiFePO4, augmentation is usually required for 10-12 years, for NMC - for 6-8 years, for AGM - replacement every 1.5-3 years (in fact, this is not an augmentation, but a complete replacement).
What is the role of the manufacturer's warranty in the LCOS calculation?
The manufacturer's Tier-1 warranty (CATL, BYD, EVE Energy, Pylontech) usually covers 10 years or 6,000 cycles of a minimum residual capacity of 70-80%. This reduces the risk premium in the discount rate and increases the banking attractiveness of the project. Noname manufacturers either do not give guarantees, or their guarantee obligations cannot be legally enforced. In fact, a Tier-1 guarantee reduces the "effective" LCOS by 10-15% due to the reduction in the cost of capital.
Conclusions of BESS Ukraine
Although LiFePO4 costs twice as much at the start (CAPEX), its cost of energy storage (LCOS) in 12 times lower. This isn't a marketing ploy -- it's the math of discounted cash flows.
For an industrial facility operating in Peak Shaving (daily cycle) or Backup mode, an investment in a quality BESS (CATL/EVE/BYD + Deye/Sungrow PCS) will pay off, and an attempt to save on technology will lead to constant losses on battery replacement.
Our recommendation: when evaluating any BESS project, ask the supplier to calculate the LCOS using the full formula of, disclosing all assumptions. If the supplier cannot or does not want to provide such a calculation, this is a red flag.
Need an audit?
Our analysts will calculate a complete financial model (CAPEX, OPEX, LCOS, ROI) for your business.