My 56 panel build...

[chrisjx]

Describing the system we installed. Would appreciate any feedback on where we might have gone terribly wrong. The system is turned off for now until I can get back and spend some time monitoring and tuning.

System Report: 56-Panel Solar Energy System – Brenham, TX​

Summary by Chris Jefferies.

---

1. Summary​

This report presents a comprehensive technical overview of a custom-designed 56-panel, 48V off-grid solar energy system located in Brenham, Texas. The system was engineered and implemented over a two-week construction period by a dedicated volunteer team. It features high-efficiency photovoltaic generation, robust energy storage, and future-ready AC load distribution infrastructure.
The primary design objective is to support an off-grid residence with a daily energy demand of approximately 60 kWh. A seasonally adjustable ground-mounted array paired with LiFePO4 battery storage and advanced inverter synchronization provides a scalable and resilient energy solution.
Site Characteristics:

• Location: Brenham, TX
• GPS: ...
• Soil Type: Dense central Texas clay
• Climate Reference: WeatherSpark Solar Data – Brenham
  • Solar irradiance: ~6.5 kWh/m²/day in summer; ~4.5 kWh/m²/day in winter
  • Wind: Moderate, 7–10 mph
  • Panel orientation: Optimized with 15° summer and 45° winter tilt

The power generation and storage systems are installed in a 20-foot container located approximately 40 feet south of the solar array, with the integrated battery, inverter, and load management components.

---

2. System Components and Configuration​

2.1 Solar Array and DC Circuitry​

• Mounting System: Sinclair Designs Sky Rack 2.0
  • Tilt range: 15° (summer) to 45° (winter)
  • Six 2' diameter, 7' deep concrete footings with 4000 psi concrete
• Photovoltaic Panels: 56x REC Alpha Pure 2 (420W each)
  • Total output: 23.5 kW DC
  • 2 rows of 28 panels each
  • Arranged in 8 strings of 7 panels (top and bottom rows)
    • Strings 1 - 4, west to east
    • Strings 5 - 8, west to east.
• Wiring:
  • 10 AWG USE-1 from string ends to junction box (~470' each of red and black wire)
  • 12 AWG THHN for underground conduit (~65', in 1.25" PVC)
  • Terminated in a 12" × 12" × 6" steel junction box with 8x 2-pole 15A, 500V DC breakers (Langir)
• String Integration: Eight PV strings feed two MPPT inputs on each of four inverters inside the container

Additional Conduit:

• A second buried PVC run (~70') was installed to support:
  • 48VDC service for a future well pump
  • 120VAC service for a water booster pump

2.2 Battery Storage and DC Bus​

• Batteries: 12x EG4 LiFePower4 48V V2 (100Ah each)
  • Total nominal capacity: 61.44 kWh
  • Configuration: Two EG4 steel racks, six batteries each
  • Cables: 2x 4 AWG per battery to rack bus rails
• Main DC Bus:
  • 600A bus bars (positive and negative), 8-post each
  • 4x 4/0 AWG cables (2 positive, 2 negative) from racks to bus bars
  • 8x 2/0 AWG cables (4 positive, 4 negative) from bus bars to inverter battery inputs
• Protection:
  • 250A fuse inline on each 4/0 positive cable (not yet covered, rear-mounted)
• Expansion Ready: Two additional EG4 LiFePower4 batteries are available for future integration

Termination: All 12 battery-to-inverter cables were terminated with hydraulic crimps and shrink-wrapped for long-term reliability and minimal voltage loss.

2.3 Inverters and AC Load Infrastructure​

• Inverters: 4x EG4 6000XP (split-phase, 6kW each)
• AC Output Wiring:
  • Each inverter terminates at a 2-pole 40A breaker in the load box
  • Wiring: 10 AWG THHN, 4-conductor (L1, L2, Neutral, Ground)
  • Neutral and ground routed to dedicated bus bars

2.4 Load Distribution Panels​

Load Box (Main Distribution Hub):

• 4x 40A breakers for inverter outputs
• 2x 100A breakers:
  • One feeds the container panel (carport loads, container lighting, HVAC)
  • One is reserved for the future residential load center

Container Panel (Secondary Panel):

• Supports lighting, receptacles, and a future 240V mini-split heat pump

2.5 Grounding​

• Ground rods installed at:
  • PV array structure and PV Junction box
  • Equipment container
  • Battery racks and inverters
  • AC and DC distribution panels

2.6 Inverter and Battery Communication​

Inverter Parallel Communication:

• Grey RS-485 cables form a closed loop among inverters:
  • Right port of inverter 1 →
  • Left port of inverter 2 and to 3→
  • Right port of inverter 4 →
  • Back to left port of inverter 1
• DIP Switch Settings:
  • Master (inverter 1): UP/UP
  • Intermediates (2, 3): DOWN/DOWN
  • Last (inverter 4): UP/UP

Battery CAN Bus Communication:

• Orange CANbus cable connects inverter P1-R485 port to master battery
• Green RS-485 daisy-chain cables link batteries from port to port
• DIP switches on batteries are configured to assign binary IDs 1–12 per manual

This configuration ensures synchronized charging/discharging, state-of-charge visibility, BMS fault monitoring, and inverter load-sharing.

---

3. Seasonal Performance Estimates​

3.1 Projected Daily Output​

Season​	Sun Hours​	Estimated Output​	Residential Load​	Surplus​
Summer (15° tilt)	6 hrs	126.9 kWh	60 kWh	+66.9 kWh
Winter (45° tilt)	4 hrs	84.6 kWh	60 kWh	+24.6 kWh

3.2 Interpretation​

• Summer: Significant energy surplus, allows room for new loads or more storage
• Winter: System still exceeds average household demand
• Seasonal tilt optimizes solar angle and total yield year-round

3.3 Initial Readings​

Initial readings were taken at approximately 6:00 PM on the day the system was powered up:

• Battery Voltage: All four inverters displayed a battery input voltage of 51.1 VDC.
• PV Input Voltage: Each inverter reported approximately 320 VDC from its connected MPPT PV inputs, with variation between strings of no more than 3–4 volts.
• There was no AC output measured.

Later, upon accessing the EG4 6000XP inverter web interfaces, (and properly configuring the inverter to battery comms) the reported data corroborated these measurements and additionally provided a battery State of Charge (SOC) of 55%. Also, the load box output measured 108VAC.
The following day at 10:30 AM, system monitoring showed:

• PV Input Voltage: Consistent with the previous day's measurements (~320 VDC)
• Battery SOC: Increased to 72%, indicating normal charging operation under morning solar conditions.
• Load voltage: 110 VAC

As of now, the system has been fully shut down, as the site will remain unattended for several months. All breakers and disconnects have been secured and verified to be in the OFF position to preserve hardware and ensure safe reactivation later.

---

4. Enhancements​

4.1 Monitoring and Diagnostics​

• Activate inverter logging and web interface
• Tune MPPTs to panel voltage maxima (Vmp ~42V)
• Monitor state-of-charge, voltage symmetry, and BMS alerts

4.2 Load Testing​

• Incrementally introduce mission-critical loads
• Confirm load startup surges do not trip inverters
• Verify thermal limits and continuous load capacity

4.3 System Improvements​

• Install DC and AC surge protection (SPD Type 2)
• Plan upgrade path from 12 to 10 AWG THHN if voltage drop is excessive
• Add thermal control to container (mini-split)

4.4 Scalability​

• Add final two batteries to bring storage to 71.7 kWh
• Add circuits for house loads (HVAC, kitchen, workshop)
• Implement automation for demand control (Home Assistant, Node-RED, Shelly relays, etc.)

---

5. Construction Notes​

Project Timeline (~ Two Weeks)​

Friday: The holes for the mounting posts were drilled.​

Monday: The posts were set in concrete.
Wednesday: The ground mount was framed up and 2 test panels mounted.
Thursday: The 56 panels were mounted to the frame.
By Sunday eve, the batteries were installed in racks, the inverters were mounted on the walls and the cables were set in place for routing and sizing.
During the course of the second week: people were traveling. We purchased overlooked and locally sourced items such as a junction box for the PV wire terminations at the array, PVC pipe (and associated breakers), the AC load box, and the Container/Carport panel.
Saturday: The trenches were dug, cables were bundled.
Sunday: The cables were pulled into PVC, and the trenches were covered.
Monday: The PV wires were terminated, 32 in the PV junction box and 16 into the 4 inverters. The AC Load box was also mounted and wired from each inverter. And finally, the container panel ,meant for provisioning power to the container and other devices in the carport, was wired from the load box. By 6PM we had all systems wired and checked. We powered up the batteries, the inverters and saw 51.1 VDC on each inverter's battery input. We could also see about 320 VDC coming in from the 8 PV strings. However we could not see any voltage on the AC output load side.
After dinner we went out to debug what seemed to be a software configuration problem. By midnight we found the issue to be that we were to plug the master inverter to master battery connection into the CAN bus socket. We plugged a lamp into the Container box outlet and achieved what we called, first-light.
System Completion Highlights:

• 23.5kW PV system online
• 12-battery LFP storage bank live
• All cable runs, protection, and grounding complete
• System ready for real-world load testing

---

References​

• Sinclair Designs Sky Rack 2.0 Install Guide (PDF)
• REC Alpha Pure 2 420W Panel Data Sheet
• EG4 6000XP Inverter Manual (PDF)
• EG4 LiFePower4 48V V2 Battery Info
• WeatherSpark Data for Brenham, TX

 

[chrisjx]

Some images of the installation.

The 56 REC Pure-2 420W panels. 8 strings of 7 panels each - 1-4 across the top, 5-8 across the bottom.


12x eg4 lifepower4 batteries, 4x EG4 6000XP inverters, then load box, then local container/carport panel.


Me; first light with a simple lamp. After I used the CANBus connection between master inverter and master battery.

 

[timselectric]

Ground rods installed at:

• PV array structure and PV Junction box
• Equipment container
• Battery racks and inverters
• AC and DC distribution panels

Oh my. That's a lot of connections to the earth, in multiple locations.
Hopefully you don't see much lighting in your area.
You didn't mention an EGC for the PV array. Hopefully it wasn't missed.

 

[niktak11]

Why so many 6000xps instead of two larger inverters?

 

[chrisjx]

Why so many 6000xps instead of two larger inverters?

I did not really evaluate 2 larger inverters. The option I considered was one large inverter like the 18k or multiple smaller inverters so if one fails we'd have to run in diminished capacity, but would still be generating.

 

[chrisjx]

Oh my. That's a lot of connections to the earth, in multiple locations.
Hopefully you don't see much lighting in your area.
You didn't mention an EGC for the PV array. Hopefully it wasn't missed.

Can you point me to a source of information on grounding a solar system. It is a subject I have not studied in detail. We have a ground mount, a metal pv junction box, a container with inverters and batteries, a covering metal carport, and of course grounding for the future house with the house's breaker panel. Are there some underlying general principles? Thanks in advance.

And yes. WE have lightening around Brenham Texas.

 

[Bluedog225]

Howdy country neighbor (Smithville). Tim’s the guy to listen to on grounding.

If you search here, you will see lots of posts on the subject. My lay person understanding is one ground rod (or two specifically spaced about 8 feet apart. (there’s NEC guidance on this) with one continuous conductor to the main panel. Then everything connected with a ground wire to the main panel.

I believe the problems are ground loops, and different charges on the multiple ground rods in the event of a lightning strike as it moves through the earth creating current in where there should be none.

Anyway, lots of ink spilled here on the topic.

And that’s a magnificent array.

 

[timselectric]

NEC article 690 for PV systems. And article 250 for general grounding.

Are there some underlying general principles?

The grounding system begins at, and is created by the N/ G bond.
The N/G bond should be located at the first means of disconnect for the AC system (usually this is the main service panel) . There should be a connection to earth at one location. This should also be connected the first means of disconnect location.
From there you run an EGC (Equipment Grounding Conductor) everywhere you run circuit conductors (along side of them).
The EGC is bonded (connected) to all conductive (metal) parts of the system.
Including (but not limited to) Equipment enclosures, junction boxes, raceways, solar panel frames and racking.
Any additional (auxiliary) ground rods (if installed) must also be bonded to the grounding system. But auxiliary ground rods should be avoided, because they create a favorable (easier than dirt) path for the gradient pulse created by a nearby lightning strike. And this surge can damage any Equipment in the path between ground rods.

This video explains it very well.

 

[chrisjx]

At about 15:30 into this video, Will talks about NOT bonding the neutral and ground and deferring to some setting (relay?) in the inverters...

Why? What would the inverters be doing?

I'm planning to bond ground (clamped to a grounding rod right outside the equipment 20' container) and neutral in the Load Box Panel. And the current thought is that it would be the "central" and only bond in my entire system from the tip of the farthest panel (~120' from container) to the well house (~70' away from container) to the house (~120' from container). Ground wires would fan out along the major routs of electrical wires and connect all the boxes and things that should be grounded. And not to DC +/-, or AC L1/L2.

Am I getting this right?

 

[timselectric]

At about 15:30 into this video, Will talks about NOT bonding the neutral and ground and deferring to some setting (relay?) in the inverters...

Why? What would the inverters be doing?

I'm planning to bond ground (clamped to a grounding rod right outside the equipment 20' container) and neutral in the Load Box Panel. And the current thought is that it would be the "central" and only bond in my entire system from the tip of the farthest panel (~120' from container) to the well house (~70' away from container) to the house (~120' from container). Ground wires would fan out along the major routs of electrical wires and connect all the boxes and things that should be grounded. And not to DC +/-, or AC L1/L2.

Am I getting this right?

You don't bond, if a bond already exists (there can be only one).
Will's system already has the bond at the grid panel.
If you don't have the grid connected to your system, then you will need to either bond, or let the inverter do it (if it has that setting).