Disclosure: This post contains affiliate links. If you purchase through our links, we may earn a small commission at no extra cost to you.
The off-grid first 30 days is not the end of the build. It is the beginning of the operation. When a customer picks up their car after a major rebuild, I spend five minutes with them in the car park before they drive away. Not because I doubt the work. Because the first 500 kilometres is when they learn the car, not just own it. They need to know what the new transmission feels like under load, what the fresh brakes sound like, what normal feels like before they can identify wrong. Your off-grid Fortress is the same. The commissioning checklist told you the system is correctly built. The burn-in test told you it holds under load. The off-grid first 30 days tells you how your system and your household actually interact: the real consumption numbers, the real recovery rate, the real autonomy at your actual load profile. That data is worth more than any design spreadsheet. Make sure your system was sized correctly before the first 30 days begins; the protocol is calibration, not correction.
Week 1: The Off-Grid First 30 Days Base Load Audit
Week 1 is the silent audit. You change nothing about your household habits. You observe. The Victron SmartShunt 500A records every Ah that leaves the battery bank and every Ah that returns. Seven days of unmodified consumption gives you the most honest load profile your system will ever produce: the real idle draw, the real peak events, the real overnight parasitic consumption.
The overnight parasitic draw is the number most new off-grid owners discover with surprise. A Cerbo GX, a router, a security camera system, a battery monitor, and a few standby electronics can add 8 to 15 Ah per day in parasitic consumption that was never included in the original load calculation. This is not a problem. It is a measurement. Once you know the number you can decide which parasitic draws are worth their consumption and which ones get switched off overnight. Week 1 gives you that number for the first time.
At the end of Week 1, log three figures: total daily Ah consumption at your normal household load, overnight parasitic draw, and peak single-day consumption. These are the three numbers that the rest of the 30-day protocol is built on. Compare them to your original design assumptions. If they are higher, you now have real data for a load management conversation. If they match, you sized the system correctly and the next three weeks are calibration, not correction. The current shunt calibration guide covers the accuracy verification that makes these Week 1 numbers trustworthy.
Week 2: The Weather Test and the Days of Autonomy Calculation
Week 2 is the weather test. You are not hoping for a cloudy week. You are waiting for the first overcast period and treating it as a planned test rather than an anxiety event. When it arrives, log four numbers each day: SOC at sunrise, Ah harvested from solar, Ah consumed by loads, and net SOC change. At the end of each day calculate your days of autonomy: remaining usable Ah in the bank divided by yesterday’s net daily consumption.
A client called me three days into their first January cloud cover in Rockwood, convinced something was wrong. The Victron Cerbo GX dashboard showed 58% SOC after 72 hours of overcast conditions. I walked them through the math: 42 Ah consumed per day at their normal load, 8 Ah recovered per day from diffuse winter light, net 34 Ah drawn from the bank each day. At 58% SOC with a 280Ah usable bank they had 4.1 days of remaining autonomy before reaching the LVC. The storm broke the next morning. The system recovered to 94% within 6 hours of clear sun. They had been within normal operating parameters the entire time. The weather test is not about surviving cloud cover. It is about understanding what the numbers mean so that cloud cover produces data instead of panic.
Days of autonomy is a more useful number than percentage SOC in isolation. A bank at 72% consuming 42 Ah per day net has 4.3 days of autonomy remaining. A bank at 72% consuming 20 Ah per day net has 9.1 days. Same percentage. Completely different management decision. Week 2 is when you first calculate this number under real conditions.
Week 3: The Second Torque Check and Thermal Verification
Week 3 is the mechanical confirmation that the first three weeks of real thermal cycling have not walked any connections. As covered in the burn-in guide, the first heat cycle produces the largest dimensional change a connection will ever experience. The second and third cycles are smaller but still present. By Week 3 the initial settling has occurred and the connections have found their stable operating dimensions.
Take the calibrated torque wrench and re-check every high-current terminal: battery terminals, busbar lugs, inverter DC input terminals, and shunt terminals. Any terminal that requires re-torquing gets logged. Run the IR thermometer scan under full load and compare every reading to the burn-in baseline from Article 148. Any Delta-T that has increased by more than 3°C since the burn-in is worth noting, even if it has not crossed the investigation threshold. The trend matters more than any single reading. After the Week 3 check, draw fresh witness marks across every bolt head and terminal face. This is the final mechanical verification before the system enters its normal seasonal maintenance rhythm covered in the off-grid system maintenance guide.
Week 4: The Energy Budget and the Load Time-Shift
Week 4 is the management calibration. You now have 21 days of real consumption data. The design assumptions are behind you. The actual numbers are in front of you.
With real daily Ah consumption confirmed, set the low voltage cutoff with confidence. You are no longer guessing at what a normal low-load week looks like. You have measured it. The LVC setting that protects the battery bank in a real Rockwood cloud event is now a calculation from real data, not an estimate from a design spreadsheet.
Identify the two or three loads that drive your peak daily consumption. For most households these are the washing machine, the dishwasher, and the hot water heating element or well pump. These loads, run during solar hours, cost zero battery capacity because the solar array is producing more than the household is consuming and the excess would go to battery charging anyway. The same loads run at 9pm cost battery capacity plus the cycling loss on the charge and discharge cycle. The shift from running loads whenever they are convenient to running high-draw loads during the 10am to 2pm solar window is the single highest-return energy management action available without adding any hardware. Week 4 is when 21 days of data makes this decision obvious rather than theoretical.
NEC and CEC: What the Electrical Codes Actually Say
NEC 690.4 requires that PV systems be installed and maintained by qualified persons throughout their operational life. The 30-day monitoring protocol is the practical implementation of that maintenance obligation for the first month of operation: it verifies that the system is performing as commissioned, identifies any thermal or mechanical changes that require correction, and produces the consumption baseline that informs every future management and maintenance decision. NEC 110.12 requires that electrical installations be maintained in a neat and workmanlike condition; the Week 3 torque check and thermal verification is the first scheduled maintenance action under this requirement.
CEC Section 64-064 requires that PV systems be commissioned and maintained in accordance with manufacturer instructions. The first 30 days of operation under the monitoring protocol described in this article represents the completion of the commissioning process as most manufacturers define it: the system has been energised, load-tested, thermally verified, and calibrated to the real consumption profile of the installation. An off-grid system that has completed the 30-day protocol and logged the results has satisfied the CEC Section 64-064 commissioning documentation requirement more completely than most residential installations ever achieve.
Quick Reference – Off-Grid First 30 Days Protocol
| Week | Focus | Key Action | Output |
|---|---|---|---|
| Week 1 | Base load audit | Log daily Ah consumption without changing habits | Real idle draw, parasitic draw, peak consumption baseline |
| Week 2 | Weather test | Track first overcast period, calculate days of autonomy | Recovery rate, autonomy at real load, cloud event calibration |
| Week 3 | Second torque check | Re-torque all terminals, IR scan vs burn-in baseline | Mechanical confirmation, updated witness marks, Delta-T trend |
| Week 4 | Energy budget | Set LVC from real data, identify loads to time-shift | Calibrated LVC, time-shift schedule, real energy budget |
| Day 30 | Log and file | Record all four weeks of data in maintenance log | First entry in the Fortress service history |
| Ongoing | Monthly SOC check | Compare monthly average SOC to Week 1 baseline | Early warning of load growth or system degradation |
On Day 30, take a screenshot of the Cerbo GX dashboard showing the 30-day consumption history and save it alongside the commissioning record and the burn-in log. This single image shows a month of real operation: the daily harvest curves, the consumption pattern, the cloud cover events, and the recovery periods. It is the most honest picture of how the system performs in the real Rockwood environment that any document can contain. File it. Date it. It is the baseline that every future annual comparison uses.
The Verdict
The off-grid first 30 days gives you three things that no amount of pre-build planning could provide: the real consumption numbers, the real recovery rate, and the calibrated confidence to manage the system without anxiety.
By Day 30:
- You know your actual daily Ah consumption, your overnight parasitic draw, and your days of autonomy at real load — numbers that belong to your specific household, your specific loads, and your specific Rockwood climate, not to a design spreadsheet
- You have completed the second torque check and thermal verification, re-drawn the witness marks, and filed the results as the first entry in the seasonal maintenance log that will run for the next decade
- You have set the LVC from real data, identified the loads worth time-shifting to solar hours, and established the energy management habits that will keep the Fortress producing reliably through year one and every year after it
In the shop we give every customer the post-delivery briefing. The first 30 days is yours. Learn the machine. Then trust it.
Questions? Drop them below.