Property-wide solar thermal polygeneration: heating, domestic hot water, air conditioning, and refrigeration from a single parabolic trough collector field, sand thermal battery, and ammonia/water absorption chiller. First physical deployment is the shipping container being built out as temporary housing (eventually a guest house). Subsequent phases extend infrastructure to the main house and shop.
Build a property-wide solar thermal polygeneration system providing heating, domestic hot water, air conditioning, and refrigeration from solar thermal input. Aggressive DIY approach. Pattern follows district heating / industrial polygeneration architecture scaled to a single property.
First physical deployment is a shipping container being built out as temporary housing that will eventually become a guest house. Subsequent phases extend infrastructure to serve the main house and shop.
¶ Design philosophy and decisions already made
- Solar thermal handles thermal loads. PV-to-resistive-heating is rejected as inefficient (20% solar-to-electric × 100% resistive vs 60-80% direct solar thermal).
- Sand thermal battery for long-duration heat storage. Sand cost ~5-20 $/kWh storage vs LFP at 150-300 $/kWh.
- Container floor gets radiant hydronic heating with thermal mass (gypcrete slab over insulation, PEX embedded). Container floor leveling required before install.
- Cooling will use ammonia/water absorption cycle. Ammonia chosen over lithium bromide because:
- Sourcing is easier (agricultural anhydrous, aqua ammonia from chemical suppliers)
- Pressure regime is more forgiving than LiBr vacuum systems
- Below-freezing capability enables both AC and refrigeration on same platform
- Ammonia smell at 5 ppm gives early leak warning far below 50 ppm OSHA limit
- Industrial knowledge base for ammonia refrigeration is mature
- Ammonia infrastructure isolated in a mechanical building. Distribution to occupied spaces uses glycol secondary loops via heat exchangers. No ammonia enters dwellings.
- Trigeneration: same collector field and primary thermal storage serves heating, DHW, and absorption cooling. Capital amortizes across all three functions.
- Tube-based parabolic trough collectors with tilt-only sun tracking (single-axis east-west tracking around a north-south axis).
- Working fluid for Phase 1: glycol/water mix (likely 40-50% propylene glycol). Targets 60-90°C output adequate for heating and DHW.
- Working fluid for later phases (driving absorption chiller): pressurized water or thermal oil to reach 130-180°C. Loop architecture should allow fluid upgrade without redesigning collectors.
- DIY fabrication of collectors including reflector surface, receiver tube design, and tracker mechanism.
¶ Sand battery decisions
- Build a large insulated vault initially. Partially fill with sand for Phase 1 to get heating online with smaller initial thermal mass. Expand sand fill as the system scales.
- Sand specifications: silica sand, dry, fine to medium grain. Cheap from local building suppliers (~50-100 $/ton delivered).
- Operating temperature range: 100-400°C target for long-term. Lower temps (100-200°C) acceptable for Phase 1 with hot water working fluid.
- Charge mechanism: embedded heat exchanger tubes carrying working fluid from collectors.
- Discharge mechanism: separate heat exchanger tubes carrying glycol/water loop to deliver heat to radiant floor and other loads.
- Insulation: high-temperature mineral wool or vermiculite, thickness sized for weeks of useful retention as design target.
- Vault construction: likely steel shell with refractory lining or fully refractory construction with steel structural support.
¶ Interim cooling and heating during phased buildout
Solar thermal heating and absorption cooling come online progressively. The container needs functional climate control immediately upon move-in, well before the absorption chiller phase is complete. Interim systems must be designed to integrate cleanly with the eventual thermal infrastructure rather than requiring removal or rework.
- Heat pump window unit (Midea, GE Profile, LG class) in the 8-10k BTU range. Heat pump preferred over resistance-heat models for backup heating efficiency in shoulder seasons.
- Mounted in the openable side of the container with proper sealing and weather flashing.
- Powered from container PV/battery system via inverter.
- Stays in place after absorption chiller comes online as backup cooling for ammonia plant service intervals, peak demand days, or absorption system failures.
- Sized for the full container cooling load so it can stand alone if needed.
¶ Interim and ongoing backup heating: propane
- Propane on-demand water heater (Rinnai, Bosch, Eccotemp class) plumbed in series with the solar buffer tank. Solar pre-heats water in buffer tank, propane unit makes up any temperature shortfall before delivery to radiant floor or DHW.
- Same propane water heater serves DHW backup year-round (showers, sinks).
- Propane storage: external tank, properly placed away from occupied space, with shutoff and pressure regulation per code.
- Plumbing integration designed in Phase 1 so propane backup is operational from day one. Solar can run "in front of" propane: solar takes whatever load it can, propane fills the gap.
- Optional secondary propane heat source: small propane space heater (Mr. Buddy class or vented wall heater) as emergency-only backup for periods when radiant floor isn't operational (commissioning, maintenance, hydronic system failure).
The propane backup is permanent infrastructure, not just an interim measure. Even after Phase 1 is fully operational, multi-day cloudy stretches in NC winter will deplete the sand battery faster than solar can recharge it. Propane fills those gaps. Designing for solar-primary, propane-backup from the start avoids ever being cold.
Progressive deployment, easier tasks first, harder tasks as understanding develops.
- Window heat pump AC for cooling and shoulder-season heating
- Propane on-demand water heater plumbed with proper integration points for future solar pre-heat
- Propane tank, regulator, and distribution
- Container insulation and air sealing complete
- Plumbing rough-in includes future solar thermal connection points
- Electrical complete for the interim systems
- Parabolic trough collector array (start small, expand)
- Single-axis tilt sun tracking
- Sand battery vault built to full eventual size, partially filled
- Buffer water tank as short-term thermal storage
- Glycol/water primary working fluid
- Radiant hydronic floor in container with gypcrete thermal mass
- Solar pre-heat plumbed in series with existing propane water heater for DHW
- Solar feeds radiant floor primary, propane feeds radiant floor as backup
- Window heat pump remains for cooling
- Ammonia/water absorption chiller, DIY build starting with experiments on used RV absorption fridge teardown, scaling to ~3-5 kW cooling capacity
- Mechanical building or isolated enclosure for ammonia plant
- Glycol secondary loop from chiller to fan coil units in container
- Heat rejection via outdoor dry cooler
- Working fluid upgrade in primary loop if needed to reach 130-160°C generator temperature for single-effect ammonia/water
- Window heat pump remains as backup cooling
- Same ammonia plant extends to lower-temperature evaporators
- Walk-in cooler and walk-in freezer (likely separate small building or attached structure)
- Replaces kitchen refrigerator with walk-in cold storage
- Distribution via ammonia piping in protected runs or via secondary cold glycol loops
- Extension to main house heating, AC, hot water
- Larger sand battery fill / additional thermal storage
- Whole-property polygeneration backbone
- Container shell stays uncut for transport (full structural strength on the road). Reinforcement tube frames for any planned window or vent openings are welded in at home, but the steel skin is left intact and cut on site after placement. Interior framing, studs, walls, and insulation built inside the existing envelope.
- Cargo doors removed permanently at the site and replaced with built insulated residential doors onto the screened porch. The container must be closable, the screened porch is unconditioned, and the radiant floor needs the box to hold heat.
- Insulation strategy needs to be addressed (likely closed-cell spray foam or rigid foam + studs + batts).
- Container floor needs leveling before radiant install (cargo containers slope to door end, cross-members create variation).
- Window AC/heater installed in the openable side as immediate climate control, stays as backup after absorption chiller comes online.
- Propane water heater installed during initial buildout with future solar pre-heat connection points already plumbed.
- Wall penetrations for thermal piping to/from mechanical area need to be planned during framing.
- Propane tank placement, line routing, and code-compliant installation handled during initial build.
- Home Assistant on Jetson Orin Nano for controls integration
- Mesh of Meshtastic LoRa nodes available for distributed telemetry
- Victron monitoring integration patterns established
- Fabrication capability: welding (TIG and MIG), 3D printing, sheet metal work, plumbing, electronics. No machining beyond drill press.
Lean into specific component recommendations, sourcing strategies, and detailed design rather than high-level concepts.
- Phase 0 interim systems specification. Window heat pump unit selection and sizing. Propane water heater selection with attention to future solar pre-heat integration (units that can accept pre-heated inlet water without problems). Propane tank sizing, location, line routing. Code compliance considerations for NC. Plumbing rough-in that anticipates Phase 1 solar tie-in without requiring teardown.
- Parabolic trough collector detailed design. Reflector geometry (focal ratio, aperture width, length per module), reflector material options for DIY at scale (polished aluminum vs aluminized mylar over substrate vs other), receiver tube design (selective coating options, glass envelope construction, end seals), mounting frame fabrication approach. Single-axis tracker mechanical design and motor/controller selection.
- Sand battery vault engineering. Vault sizing for various fill levels. Insulation specification (material choices, thickness for target retention duration). Heat exchanger tube layout inside sand (charge and discharge circuits, sizing, material). Structural design for thermal cycling. Charging temperature targets vs cost tradeoff.
- Working fluid selection and loop design for Phase 1. Glycol concentration, expansion tank sizing, pressure relief and overheating protection (heat dump strategy when summer generation exceeds demand), drainback vs pressurized closed-loop architecture, pump sizing and selection (12V DC pumps preferred).
- Container radiant floor design. Gypcrete vs alternative thermal mass approaches given container floor structural capacity. PEX layout. Manifold and zone valve sizing. Temperature control strategy for high thermal inertia slab.
- Buffer tank specification and propane integration. Size and type for Phase 1. Stratification design. Heat exchanger coils (collector loop and load loop) inside the tank. Insulation. Plumbing topology for solar buffer tank feeding propane on-demand heater for DHW. Integration with sand battery (parallel storage, or series with buffer tank as short-term and sand as long-term).
- Controls architecture. Integration with existing Home Assistant on Jetson Orin Nano. Sensor selection (temperature, flow, pressure). ESP32 or similar microcontroller-based field controllers. Differential controller logic for collector pump. Predictive controls using weather forecast data. Tracker control firmware. Smart switching between solar primary and propane backup based on buffer tank temperature.
- Phase 2 ammonia/water absorption chiller design. Starting with RV absorption fridge teardown for learning, scaling to ~3-5 kW cooling capacity for container. Component-level design: generator, rectifier, condenser, evaporator, absorber, solution heat exchanger, expansion valves. Pressure vessel design and safety. Ammonia sourcing and handling.
- Phase 3 cold storage extension. Multiple-evaporator architecture on single ammonia plant. Walk-in cooler/freezer construction. Distribution piping (ammonia in protected runs vs secondary glycol loops). Insulation specifications for cold storage structures.
- System monitoring and instrumentation. What to measure for performance tracking, troubleshooting, and long-term optimization. Logging into existing time-series infrastructure.
- Realistic sequencing and dependencies. What has to be done before what. Critical path items. Long-lead components to order early. Things that can be deferred until later phases without rework.
¶ Constraints and preferences
- DIY-first approach for fabrication, sourcing, and assembly. Off-the-shelf only when DIY doesn't make sense.
- All thermal infrastructure should be designed to scale to whole-property use eventually, not just container-scale.
- Safety-first on ammonia handling (mechanical room isolation, ventilation, detection, proper PPE).
- Designs should accommodate future expansion (larger sand fill, more collectors, additional buildings) without requiring rework.
- Interim systems (window AC, propane heat) must integrate cleanly with eventual solar thermal infrastructure and remain useful as backup after solar comes online.
- Avoid em dashes, rhetorical triads, corporate jargon in any written output.
- The user pushes back immediately when something is wrong, prefers honest tradeoffs over sugarcoated decisions.
Container interior framing not yet built. Phase 0 interim systems and solar thermal Phase 1 infrastructure can be planned and built in parallel. The propane and window AC go in with the container buildout, while the parabolic collectors, sand battery, and buffer tank work proceeds outside the container in the mechanical area.
First concrete output should be a detailed parts list, sourcing plan, and rough cost estimate for Phase 0 (interim systems with future-integration plumbing) and Phase 1 (heating + DHW + sand battery foundation) so the user can begin acquisition and fabrication.
¶ Decisions added 2026-05-21 (container build and plant structure)
These refine the container deployment and the plant architecture. Full container and cluster build detail lives in Container Home Base.
Mechanical plant is its own structure. The collector feed, sand battery, buffer tank, ammonia chiller, and pumps live in a dedicated mechanical building, not attached to any dwelling. Distribution to occupied spaces is glycol secondary loops through heat exchangers. Refinements:
- Minimize heat-exchanger stages on the high-temp path that drives the ammonia generator (130-180 C). Each HX costs 5-10 C of approach, free on the low-temp floor side but expensive on the generator side.
- Give the ammonia its own ventilated room with leak detection, separated from the sand battery and any hot surfaces. Ammonia is flammable at 15-28% in air and should not be able to reach a 400 C battery wall.
- Use pre-insulated buried distribution line and keep runs short. Plant siting is a three-way compromise: close to the buildings served, unobstructed sun for the adjacent trough field, and safe isolation.
- Size the vault, structure, and distribution stubs for the eventual main house and shop now, not just the container.
- Brazed-plate exchangers (off-the-shelf) for liquid-to-liquid transfers are a buy. The DIY heat exchangers are the charge and discharge tube bundles buried in the sand.
Buy or hire the lethal 10%. The ammonia loop is the one subsystem to outsource: ASME-stamped pressure vessels from a certified shop for the generator, absorber, and condenser, and an ammonia refrigeration contractor (IIAR-familiar) for charging and commissioning. Everything else (troughs, reflectors, tracker, sand vault, low-temp glycol loops, radiant floor, controls) stays DIY. The ammonia choice is what creates that hazard zone, and it is the price of the Phase 3 walk-in freezer capability. The mechanical-room isolation plus glycol secondary loops already match the ASHRAE 15 code architecture.
Radiant floor transport handling. Rough in the PEX at home and pressure-test it, then keep the loop air-charged (not water) through the move so the gauge doubles as a transit-damage and slab-pour puncture detector. Pour the gypcrete only after the container is set on its final piers. A rigid thermal-mass slab cracks on the road and adds thousands of pounds to the haul.