Battery Bank Autonomy and Weak Block in an Off-Grid Telecom Repeater
A remote telecom repeater runs on a 24 V DC bus with a continuous load of 3.5 A, giving about 84 W and 2.02 kWh per day. The battery bank has eight 12 V, 150 Ah VRLA blocks arranged as four parallel strings of two blocks in series, giving 24 V and 600 Ah nominal? No: the installed bank is only two parallel strings of two blocks, giving 24 V and 300 Ah. A 1.98 kWp array of six 330 Wp modules feeds a 60 A MPPT charge controller. The winter peak sun hours are 4.0 and the system derating is 0.75. Absorption is set to 28.8 V and float to 27.2 V, but the controller logs only about 20 minutes of absorption before switching to float. After a cloudy spell the load disconnects at 22.0 V. One block in string A is warm during charging and reads 0.3 V lower than its neighbour. The operator wants to know whether the array, the battery bank, or the controller is the main problem.
1. Calculate the present autonomy and the battery capacity needed for three days of autonomy at 50% depth of discharge. State whether the installed bank is adequate.
Read the worked answer
Daily load
Energy = 24 V × 3.5 A × 24 h = 2016 Wh = 2.02 kWh/day.
Present bank
Nominal energy = 24 V × 300 Ah = 7.2 kWh.
Usable at 50% DoD = 3.6 kWh.
Autonomy = 3.6 / 2.02 = 1.78 days, less than two days.
Required for three days
Usable energy needed = 2.02 × 3 = 6.05 kWh.
Nominal energy at 50% DoD = 6.05 / 0.5 = 12.1 kWh.
Capacity at 24 V = 12.1 × 1000 / 24 = 504 Ah.
Adding about 15% for ageing and temperature gives roughly 580-600 Ah. The installed 300 Ah bank is about half the required capacity. The main problem is the battery bank, not the array: the array can produce about 1.98 × 4.0 × 0.75 = 5.94 kWh/day, which is well above the 2.02 kWh load even after battery round-trip losses.
2. Interpret the warm block and the 0.3 V difference. What tests would you perform, and what replacement rule would you apply?
Read the worked answer
A block that is warm during charge and sits 0.3 V below its neighbour is showing a probable internal short, advanced sulphation, or a shorted cell. In a series string, that block limits the whole string, causes the controller to reach absorption voltage early and then fall to float, and can overheat or gas.
Tests:
- Measure each block’s open-circuit voltage at rest, after a short settling period.
- Measure block voltages under charge and under a known load; the weak block will fall faster.
- Conduct a conductance or internal-resistance test and compare all blocks.
- If the blocks are flooded, check specific gravity; VRLA blocks cannot be checked this way.
- Check terminal torque, corrosion, and whether the warm block is simply a loose connection.
Do not simply replace one block in an old bank unless it is an exact match in make, model, capacity, and age. A new block in series with aged blocks will be overcharged or undercharged and will fail early. Replace the affected string, or preferably the whole bank with matched blocks. If one block is replaced, follow the manufacturer’s equalisation instructions; VRLA types usually do not permit equalisation. After replacement, record the block voltages and recheck the controller settings.
3. Recommend charge-controller settings and checks for the enlarged bank. Include the risk of the 60 A controller and the observed short absorption time.
Read the worked answer
For a 24 V VRLA bank, typical settings at 25 °C are absorption about 28.8 V (2.40 V/cell) and float about 27.2 V (2.27 V/cell), with temperature compensation of roughly -60 mV/°C for the 24 V bank. Confirm against the battery manufacturer’s data.
Absorption should last 2-3 hours, or until the charge current falls to about 1% of capacity. For a 600 Ah bank that is about 6 A. The observed 20-minute absorption is too short; the controller may be reaching absorption voltage while the bank is still partially discharged, then dropping to float. Set the absorption time and finish-current condition properly, and check that the controller’s temperature sensor is fitted to the battery.
Check the 60 A controller against the array. A 1.98 kWp array at 24 V can theoretically deliver about 82 A, so the 60 A controller will clip on clear cold days. That is acceptable only if the manufacturer allows it and the array is not needed for full recovery; otherwise use a larger controller. At 600 Ah, C/10 is 60 A, so the controller is at the lower edge of recommended charge current. Verify cable voltage drop, controller derating in high ambient temperature, and that the array winter yield of about 5.9 kWh/day still covers the load after ageing and dirty modules.