10 Off-Grid Solar Mistakes (and How to Avoid Them)
The 10 most costly off-grid solar mistakes, invisible at commissioning but expensive in service. Each one includes a diagnostic test and corrective action.
The most expensive off-grid solar mistakes are invisible at commissioning. The system energises, the solar array produces positive yield on the Cerbo GX display, the loads run without interruption, and the client signs off. The mistake is already in the installation. It is producing its effect silently, a battery cycling at ninety percent depth of discharge because the bank was sized to nameplate capacity without applying the DoD factor, a cable running above its derated thermal limit because the derating calculation was skipped, a BMS transmitting CVL and CCL that the charge sources cannot receive because the VE.Can connector was never fully seated.
Twelve to eighteen months later the client calls. The battery no longer reaches its rated autonomy. The cable insulation is discoloured and soft to the touch. The inverter trips every time the air conditioner starts. The client’s first instinct is to blame the product. The product is functioning correctly. The mistake was made at the design table, at the installation stage, or during the thirty minutes of commissioning that was compressed into ten to get to the next job.
This post documents the ten recurring mistakes that produce these callbacks, organised by the stage at which they occur, with the root cause, the symptom, the diagnostic measurement, and the corrective action for each.
Design-Stage Mistakes
Design-stage mistakes are the most consequential category because every downstream calculation inherits their errors. A load audit that overstates daily demand by 30 percent produces an oversized array and battery bank. A load audit that omits the overnight standby loads understates daily demand by 15 to 25 percent and produces a battery bank that runs short before sunrise. Neither error is visible at commissioning because the system produces positive yield and the loads run, and neither error is self-correcting without someone recalculating the design from measured data.
Mistake 1: Load Audit Using Nameplate Wattage Instead of Running Power
Every electrical appliance carries a nameplate rating that is the maximum power the device can draw under worst-case conditions. For resistive loads such as LED lighting, the nameplate rating is close to the actual running power. For motor loads such as air conditioners, refrigerators, and water pumps, the nameplate rating is significantly above the running power because the motor draws its nameplate current only during starting and at full mechanical load. An air conditioner with a nameplate rating of 1,800W may draw only 1,200W at its typical part-load operating point in a well-insulated room at steady-state temperature. Using 1,800W in the load audit overstates the daily energy contribution of that load by 50 percent.
The corrective approach is to measure running power with a clamp meter or plug-in power meter during normal operating conditions, not at startup. The measured running power, not the nameplate rating, is the figure that belongs in the load audit.
Diagnostic: if the Cerbo GX daily consumption is consistently 20 to 30 percent below the load audit estimate after the first month of operation, nameplate wattage was used where running power should have been measured. Refer to how to do a proper load audit before sizing an off-grid system for the correct methodology.
Mistake 2: Omitting Phantom and Standby Loads
Phantom loads are the power consumed by appliances in standby mode, and they are omitted from the majority of residential load audits because they are invisible during a visual inspection of the property. A satellite decoder that draws 18W in standby mode, a WiFi router that draws 12W continuously, a microwave with a clock display that draws 3W, and an inverter’s own standby consumption of 15W combine to a phantom load of 48W. Over 24 hours, this phantom load consumes 1,152Wh, more than the daily energy demand of the entire lighting circuit. On a battery bank sized for two autonomy days, the unaccounted phantom load reduces the effective autonomy by 15 to 25 percent.
Diagnostic: the Cerbo GX’s AC load reading at 02:00 in the VRM historical data shows the sum of all phantom loads active overnight. A non-zero reading at that hour that exceeds 50W on a residential system indicates unaccounted standby loads. Refer to how to account for phantom loads and standby power in off-grid energy budgets for the identification and quantification methodology.
Mistake 3: Battery Bank Sized to Nameplate Without DoD Factor
A 14.2kWh battery bank does not deliver 14.2kWh of usable energy. It delivers 12.78kWh at 90 percent depth of discharge. An installer who uses the 14.2kWh nameplate figure directly in the autonomy calculation produces a bank that delivers 10 percent less energy than the design assumes. The error compounds when the installer also omits the autonomy day multiplier and sizes the bank only to a single day’s demand. A bank sized for one day’s demand at nameplate capacity has an effective usable capacity of 0.90 days at the design load. On a day with 60 percent solar harvest, a common overcast condition in coastal West Africa, the system will reach the BMS low voltage cutoff before midnight.
Diagnostic: if the system shuts down on low battery voltage before the expected end of autonomy on overcast days, calculate the actual usable capacity at the installed DoD and compare against the daily demand multiplied by the design autonomy days. Refer to battery bank sizing for off-grid systems: capacity, BMS selection, and cycle life for the three-variable sizing framework.
Mistake 4: Inverter Sized to Average Load Not Peak Simultaneous Load
The inverter’s continuous rating must exceed the peak simultaneous load. The inverter’s surge rating must exceed the peak simultaneous load plus the starting inrush current of the largest motor load that could start while the other loads are running. The failure pattern is consistent across dozens of post-commissioning callbacks: the system operates correctly for weeks or months under the typical daily load combination. A motor load that was not identified during the load audit starts during a high background load condition. The combined surge demand exceeds the inverter’s surge rating. The inverter shuts down on overload. The client reports that the system has stopped working.
Installation-stage mistakes are made correctly on paper and incorrectly in the field. The cable cross-section specified in the design is correct. The cable installed in the wall is one size smaller because the specified size was not available at the supplier on installation day and the substitution was not documented. The earth bond is shown correctly on the design drawing as a single point at the inverter. A second bond was added at the battery enclosure during installation because the installer believed that any metal enclosure should be earthed. The system operates correctly in both cases — until it does not.
Mistake 5: DC Cable Undersizing: Thermal Derating Not Applied
The most dangerous characteristic of an undersized DC cable is that it produces no immediate fault. A cable carrying 20 percent more current than its derated thermal limit does not trip a circuit breaker or blow a fuse if the fuse was also incorrectly rated. The cable heats. The insulation softens. Over months and years the insulation becomes brittle, cracks, and eventually fails, either as a ground fault, a short circuit, or a fire.
The derating factors that must be applied to every DC cable rating are installation method, bundling, and ambient temperature. A 35mm² cable with a free-air nameplate rating of 158A carries only 82A when installed in a conduit bundle of three cables at 50 degrees Celsius ambient, a 48 percent reduction in capacity.
Diagnostic: calculate the derated current capacity of the installed cable using the three-factor derating equation. Compare against the maximum circuit current. If the maximum circuit current exceeds the derated capacity, the cable is undersized. Refer to DC cable sizing for off-grid solar systems: the two-constraint framework for the full derating methodology and reference tables.
Mistake 6: Fuse Rated to Load Not to Cable
The fuse protects the cable, not the load. A fuse selected to match the inverter’s rated input current rather than the cable’s derated current rating does not protect the cable from thermal stress. If the cable’s derated current rating is 82A and the fuse is rated at 100A to match the inverter’s rated input current, the fuse will not trip on a sustained cable overload between 82A and 100A.
The cable will overheat at any current above 82A without any protection device responding. The correct fuse rating satisfies two conditions: greater than the maximum continuous circuit current so the fuse does not trip during normal operation, and at or below the cable’s derated current rating so the fuse trips before the cable is thermally damaged.
Diagnostic: check the fuse rating against the cable’s derated current rating, not the load current. If the fuse rating exceeds the cable’s derated current rating, the cable is unprotected for any current between those two values. Refer to DC cable sizing for off-grid solar systems: the two-constraint framework for the fusing rule and the correct rating methodology.
Mistake 7: Multiple Earth Bond Points on DC Negative Bus
The DC negative bus must be bonded to the system protective earth at exactly one point. When a second earth bond is added at the battery enclosure, any voltage difference between the two bond points drives a current through the earth conductor between them.
This circulating earth current is uncontrolled, unfused, and flows through whatever conductive path is available including metal enclosure surfaces, cable armour, and conduit. At low levels it produces electrochemical corrosion at every earth connection point. At higher levels it produces interference in BMS and monitoring communication circuits, causing erratic BMS behaviour, false alarms, and nuisance RCD tripping when the generator is connected. These symptoms develop over months, not hours, and are almost never attributed to the earth bonding error until they are severe enough to investigate.
Commissioning-stage mistakes are made during the final hour of an installation when the pressure to energise, demonstrate, and hand over is highest and the attention to systematic verification is lowest. They are also the easiest mistakes to prevent, because the commissioning sequence in off-grid solar system commissioning and troubleshooting: the complete field guide is a checklist that takes 90 minutes to complete correctly and prevents every mistake in this section.
Mistake 8: BMS Communication Not Verified Before Activating Charge Sources
The BMS communication link between the Pylontech battery bank and the Cerbo GX is the mechanism that allows the charge sources to serve the battery rather than the battery protecting itself from the charge sources. When the CVL, CCL, and DCL are absent — because the VE.Can connector was not fully seated, the CAN termination resistor is incorrectly placed, or the VE.Can-to-CAN adapter was not connected — the charge sources revert to their locally programmed charge profiles and operate without any BMS guidance.
The more dangerous scenario is invisible: a locally programmed profile that is slightly above the BMS’s intended CVL, producing a battery that is consistently overcharged by 0.2 to 0.4V on every cycle. At this level the overcharge is not sufficient to trigger the BMS hardware disconnect but is sufficient to accelerate SEI layer growth on the anode, reducing the battery’s cycle life by 20 to 40 percent compared to a correctly communicating system. The battery performs normally for 18 months and then begins losing capacity faster than the manufacturer’s specification predicts. The client disputes the warranty. The BMS communication log in the VRM data shows that communication was never established.
Diagnostic: before activating any charge source, navigate to the Cerbo GX device list and confirm the Pylontech BMS appears as a connected device with CVL 57.6V, CCL 100A, and DCL 148A displayed. If any of these values are absent or incorrect, do not activate the charge sources. Refer to off-grid solar system commissioning and troubleshooting: the complete field guide for the full commissioning sequence and BMS communication verification procedure.
Symptom
Likely Cause
First Check
Fix
BMS not on Cerbo GX device list
Unseated VE.Can connector
Reseat VE.Can connector on Cerbo GX rear panel
Reseat and verify CVL/CCL/DCL appear
BMS on list but no CVL/CCL/DCL
Incorrect CAN termination
Check termination resistor at one end only
Remove second termination resistor
BMS appears then disappears
Intermittent CAN cable
Test cable continuity
Replace CAN cable
CVL/CCL/DCL wrong values
Wrong BMS protocol selected
Verify Pylontech protocol in Cerbo GX settings
Select correct protocol
Comms fails after firmware update
Firmware incompatibility
Check Victron supported battery list
Roll back or update firmware
Mistake 9: Commissioning Documentation Not Produced
Commissioning documentation is not a bureaucratic requirement. It is the only record that allows a fault occurring 18 months after installation to be diagnosed in minutes rather than hours. Without a documented CVL, CCL, and DCL from commissioning, there is no way to confirm whether the BMS communication was established on day one. Without a documented AC output voltage measurement, there is no baseline for determining whether the inverter’s output has drifted. Without an as-built wiring diagram showing the actual cable cross-sections installed, every cable in the system must be physically traced and measured before the thermal derating calculation can be verified.
The specific consequence most commonly encountered in the field is the warranty dispute. A battery bank that loses 25 percent of its capacity within 18 months of installation may be a defective product or it may be the result of 18 months of operation without BMS communication. The manufacturer’s warranty team will request the commissioning records. An installer who cannot produce commissioning records documenting that BMS communication was established and verified on day one has no defence against a warranty denial on grounds of improper installation.
Diagnostic: check whether the five required documents exist — as-built wiring diagram, commissioning test results, Cerbo GX configuration export, VRM portal credentials, and serial numbers and warranty registrations. Refer to off-grid solar system commissioning and troubleshooting: the complete field guide for the complete documentation requirements.
Post-Commissioning Mistakes
Post-commissioning mistakes are made after the installer has left and the system has been handed over to the client. They are the most expensive category in terms of repair cost per incident because they frequently involve purchasing components that are incompatible with the existing system, installing them, discovering the incompatibility, and then either returning the component or operating the system outside its design parameters.
Mistake 10: System Expansion Without Constraint Analysis
The system expansion constraint analysis in off-grid solar system expansion without triggering failures: the complete constraint analysis identifies five constraints that every expansion must be checked against before any component is purchased: the MPPT controller input voltage, the MPPT controller output current, the inverter continuous and surge rating, the battery charge and discharge current limits, and the DC cable ratings. A client or third-party contractor who adds a load, a panel, or a battery unit without performing this analysis will discover the binding constraint at the worst possible moment.
The three most common expansion errors each produce a characteristic failure mode. Adding panels to an existing string without recalculating the cold Voc produces a string that exceeds the MPPT controller’s input voltage limit on the first cold morning and triggers a high voltage alarm. The client reports the MPPT controller has failed.
The controller has not failed, it has correctly protected itself from an overvoltage condition produced by the expansion. Adding a fifth Pylontech unit to a four-unit bank without recognising the four-string BMS limit produces a system where the fifth unit is either unmanaged by any BMS or managed by a second BMS whose CVL and CCL may conflict with the first. Adding a new load without checking the inverter’s remaining surge capacity produces the overload shutdown described in Mistake 4, except that this time the load was added after commissioning.
Each of these three expansion errors would have been caught by running the constraint analysis before the component was purchased. The constraint analysis takes 30 minutes with the commissioning documentation as the reference. Without commissioning documentation, it requires a site visit to measure and inspect the existing installation, which is why Mistake 9 and Mistake 10 compound each other.
Every mistake in this post has a diagnostic test that confirms or eliminates it as the root cause without requiring component replacement. The consistent failure of field diagnosis is not the absence of knowledge about what the root cause might be, most experienced installers can list the likely causes of a given symptom. The failure is the absence of a structured approach that takes one measurement, interprets it against the commissioning baseline, and either confirms the most likely cause or eliminates it and moves to the next candidate.
The four-step diagnostic framework that applies to every fault on every system is: confirm the symptom precisely, identify which subsystem the symptom implicates, take the one measurement that most directly addresses the most likely root cause, and eliminate causes systematically before replacing any component.
The commissioning documentation is the baseline that makes this framework operational. A documented CVL of 57.6V on day one, compared against a current CVL reading of zero, confirms that BMS communication has been lost since commissioning. A documented AC output voltage of 230.4V on day one, compared against a current reading of 214V, confirms that the inverter’s output has drifted and is not a design error. Without the baseline, both conditions require a judgement call rather than a measurement comparison.
The mistake most resistant to this framework is Mistake 7, multiple earth bond points, because the measurement that confirms it, a non-zero DC negative to earth voltage, is not part of any standard fault diagnosis checklist. An installer who finds intermittent BMS communication loss, nuisance RCD tripping, and unexplained corrosion at earth connections on the same system should measure the DC negative to earth voltage before investigating any other cause. If the reading is non-zero, the earth bonding architecture is the root cause of all three symptoms simultaneously.
Diagnostic Framework Reference:
Step 1 -> confirm the symptom precisely: what exactly is happening, when, and under what conditions
Step 2 -> identify the implicated subsystem: which component or circuit does this symptom point to
Step 3 -> take one measurement: the single measurement most directly addressing the most likely cause
Step 4 -> eliminate systematically: confirm or rule out each candidate before replacing any component
Baseline -> all measurements compared against commissioning documentation values
Rule -> never replace a component before a measurement has confirmed it is the fault source
The ten mistakes in this post are consolidated in the reference table below. Each row provides the mistake, the stage, the symptom that surfaces in service, the first measurement that confirms or eliminates it, the likely root cause, the corrective action, and the post where the correct methodology is documented.
Why does my off-grid system shut down unexpectedly?
The three most common causes are an inverter overload from a motor load starting during a high background load condition, a battery reaching its BMS low voltage cutoff before the expected end of autonomy because the bank was undersized or phantom loads were not accounted for, and a BMS communication failure causing the charge sources to operate on incorrect profiles that force the BMS into a hardware disconnect. The Cerbo GX alarm log identifies which condition occurred and at what time. Start there before replacing any component.
How do I know if my battery bank is undersized?
If the system reaches the BMS low voltage cutoff before the expected end of the design autonomy period on a day with normal solar harvest, the battery bank is either undersized or is not delivering its rated usable capacity due to degradation. Calculate the actual usable capacity at the installed depth of discharge and compare against the daily energy demand multiplied by the design autonomy days. Refer to battery bank sizing for off-grid systems: capacity, BMS selection, and cycle life for the three-variable sizing framework.
What is the most common cause of BMS communication failure?
An unseated VE.Can connector on the Cerbo GX rear panel accounts for the majority of BMS communication failures in the field. The second most common cause is incorrect CAN termination, two termination resistors installed at both ends of the CAN bus rather than one at each end. Both faults are detectable by physical inspection in under five minutes. Refer to off-grid solar system commissioning and troubleshooting: the complete field guide for the full BMS communication diagnostic pathway.
Why is my generator not charging my battery?
The four most common causes are the AC input current limit not configured, the generator’s output voltage sagging below the Multiplus-II’s stable operation threshold under combined load, BMS communication failure causing the charge sources to receive no CCL, and the minimum run time not set causing the generator to stop before the BMS stop condition is reached. Measure the generator output voltage at the Multiplus-II AC input terminals under combined load before investigating any other cause. Refer to generator integration, sizing, and hybrid operation for off-grid solar systems for the full generator integration framework.
What is the most common off-grid solar mistake in Nigeria?
The most common mistake is sizing the battery bank using nameplate capacity without applying depth of discharge and C-rate derating. A battery bank sized to match daily load at 100% nameplate capacity delivers only 50–80% of that energy in real operation, the rest is either unusable due to chemistry limits or accelerates premature degradation. The second most common mistake is configuring the inverter with lead-acid voltage settings when a LiFePO4 battery is installed, which silently degrades the battery on every charge cycle.
Why do most off-grid solar systems fail within 2 years?
Most early failures trace back to three compounding mistakes made at design and commissioning. First, the battery bank is undersized relative to the real load including phantom loads and derating. Second, the inverter is not configured correctly for the battery chemistry. Third, no monitoring is set up, so the degradation caused by the first two mistakes accumulates silently until a component failure forces the first service call. By that point the battery has often been damaged beyond economic repair.
What happens if you connect a lithium battery to an inverter set for lead-acid?
The inverter applies an absorption voltage between 57.6V and 59.2V on a 48V system higher than the 56.8V maximum recommended for LiFePO4. Every charge cycle pushes the cells into mild overvoltage. The BMS may or may not intervene depending on its configuration. Over months this accelerates SEI layer growth on the anode and cathode stress, progressively reducing capacity. The system appears to work normally, the battery charges and discharges but the damage accumulates invisibly until capacity drops noticeably, typically at 12 to 24 months.
How do I know if my off-grid system was commissioned correctly?
A correctly commissioned system has five documents at handover: an as-built wiring diagram, a commissioning test results sheet with measured values, a Cerbo GX or equivalent monitoring platform configuration export, VRM portal credentials with the system already online, and serial numbers and warranty registrations for all major components. If any of these five documents are missing, the commissioning was not completed correctly. The absence of documentation is not a paperwork issue, it means the systematic verification was not done.
Can I expand an off-grid solar system after it has been installed?
Yes, but only after running a five-constraint check before purchasing any new component. The constraints are: inverter continuous and surge rating headroom for the additional load; battery bank capacity relative to the new daily energy demand; MPPT controller input voltage and output current limits for an expanded array; DC cable ampacity for the increased currents; and BMS charge and discharge current limits for a larger battery bank. Skipping this check and connecting a new component, a second air conditioner, an additional battery, a larger array is how expansion creates failures in previously working systems.
What diagnostic tool is most useful for off-grid solar fault finding?
For Victron-based systems, the VRM portal is the most powerful diagnostic tool available. It stores 90 days of historical system data including cell voltages, SOC, charge and discharge current, inverter load, and all alarms. Most faults that appear sudden in the field were preceded by warning signs in the VRM data days or weeks earlier. For systems without remote monitoring, a clamp meter and a multimeter are the two essential tools, the clamp meter for measuring actual current versus expected current, and the multimeter for voltage verification at each connection point in the diagnostic sequence.
Conclusion
The ten mistakes in this post are not rare edge cases or the result of unusual circumstances. They are the recurring failure patterns from off-grid installations across the residential and small commercial market, documented from callbacks, warranty disputes, and fault diagnoses on systems that were installed without the methodologies that each referenced post establishes.
Every mistake has a prevention. The installer who applies the load audit methodology, the cable sizing framework, the commissioning sequence, and the expansion constraint analysis will not produce any of the ten mistakes in this post. The methodology is documented. The prevention is a decision to apply it.
I am Engr. Ubokobong Ekpenyong, a solar specialist and lithium battery systems engineer with over five years of hands-on experience designing, assembling, and commissioning off-grid solar and energy storage systems. My work focuses on lithium battery pack architecture, BMS configuration, and system reliability in off-grid and high-demand environments.
Contains information related to marketing campaigns of the user. These are shared with Google AdWords / Google Ads when the Google Ads and Google Analytics accounts are linked together.
90 days
__utma
ID used to identify users and sessions
2 years after last activity
__utmt
Used to monitor number of Google Analytics server requests
10 minutes
__utmb
Used to distinguish new sessions and visits. This cookie is set when the GA.js javascript library is loaded and there is no existing __utmb cookie. The cookie is updated every time data is sent to the Google Analytics server.
30 minutes after last activity
__utmc
Used only with old Urchin versions of Google Analytics and not with GA.js. Was used to distinguish between new sessions and visits at the end of a session.
End of session (browser)
__utmz
Contains information about the traffic source or campaign that directed user to the website. The cookie is set when the GA.js javascript is loaded and updated when data is sent to the Google Anaytics server
6 months after last activity
__utmv
Contains custom information set by the web developer via the _setCustomVar method in Google Analytics. This cookie is updated every time new data is sent to the Google Analytics server.
2 years after last activity
__utmx
Used to determine whether a user is included in an A / B or Multivariate test.
18 months
_ga
ID used to identify users
2 years
_gali
Used by Google Analytics to determine which links on a page are being clicked
30 seconds
_ga_
ID used to identify users
2 years
_gid
ID used to identify users for 24 hours after last activity
24 hours
_gat
Used to monitor number of Google Analytics server requests when using Google Tag Manager
1 minute
You can find more information in our Cookie Policy and .
