01At a glance
Design-point figures from the specification and turbine data sheet, with what the two delivered sets actually demonstrated.
02Purpose of the silent 1.5 kW design
Why the Army wanted it, and what "silent power plant" meant in the specification.
The set was developed by Sundstrand Aviation (Rockford, Illinois) for the U.S. Army Mobility Equipment Research and Development Center at Fort Belvoir under contract DAAK02-72-C-0472. Its job was to replace small, noisy field engine-generators with a tactical power source that an enemy could not hear: the governing requirement was that the running set be inaudible at 100 metres.
A closed Rankine cycle was chosen because it has no reciprocating parts and no internal combustion pulses. Fuel burns continuously in a quiet "blue-flame" recirculating combustor, the heat boils an organic working fluid, and a single high-speed turbine spins a brushless alternator. Exhaust and cooling air leave together under a sound-lined weather cap.
The 28 V DC output matched U.S. military vehicle, radio and battery standards, so the set could power command posts, communications and sensors, and charge batteries, from any fuel in the supply chain.
Key specification requirements (Table III-A)
- 1.5 kW, 28 V DC, closed Rankine cycle; portable, skid-mounted
- Silent at 100 m
- Multi-fuel (demonstrated on MIL-T-5161 and W-F-800 fuels)
- −65 °F to +125 °F, any humidity; rain 12 in/h, wind 40 mph
- ≤ 150 lb dry, ≤ 8 ft³, ≥ 7.5 % set efficiency at 1.5 kW
- ≥ 3,000 h life; ≥ 95 % reliability; 31° inclined operation
- Start within 10 min from battery, external power or hand crank
- Survive 18 in flat / 12 in end drops, rail impact at 10 mph
- Hermetic working-fluid loop; no significant fluid loss in 10,000 h
03The four source documents
What each attached PDF contributes.
Sundstrand ATR 1182 — Final Report, Set No. 1 (May 1974)
Full design, analysis, fabrication and test record for the first set (June 1972 – Dec 1973): working-fluid selection, CRU design, heater/combustor, regenerator and condenser, accessory drives, power conditioning, shock/noise analysis, and 49 hot tests. Identifies three blocking problems — low turbine performance (nozzles spaced too far apart, giving 48 % of theoretical turbine efficiency), turbine-exhaust quenching from pitot-pump leakage, and CRU structural flexibility causing light rubs. Set 1 was delivered; Set 2 was held for a new nozzle plate.
Sundstrand AER 1263 — Program Report, Set No. 2 (Sept 1975)
Documents the improved second set: close-spaced nozzle plate, stiffer CRU with larger shaft and the pitot pump moved aft, start pump replacing the accumulator start, non-sticking control valves, about 76 W lower parasitic power, and gearbox noise work. Set 2 was self-sufficient except for battery start and reached 477 W net, held back mainly by low pitot-pump efficiency (12–17 % vs 35 % design), heat shunts and lower heater/regenerator performance. Projects 1.2–1.6 kW at 10.4–13.9 % with the listed fixes.
Sundstrand 1.5 kW ORC Turbine Wheel (Nov 2026)
A geometric reconstruction of the axial impulse wheel from the ATR 1182 turbine data sheet: 3.000 in pitch diameter, 111 blades, 24° blade angles, 7 of 10 converging-diverging nozzles active, Mach 3.36 jet, U/C₀ = 0.362, 61 % design efficiency. Cross-checks geometry, Ns·Ds and velocity triangles, and flags an inconsistency between the listed 0.033 lb/s flow and 3.43 hp output (≈0.050 lb/s would be consistent).
NASA TM X-2200 — Motor-Starting Characteristics of a Modified Lundell Alternator (March 1971)
Lewis Research Center tests of a 10.7 kW, 36,000 rpm solid-rotor Lundell (Rice) alternator from the Brayton space-power program, started as an induction motor. Relevant because the 1.5 kW set uses the same brushless Lundell alternator family on a single high-speed shaft, and motor-starting through the alternator is a candidate way to bootstrap such a turbo-alternator.
04How the set works
A supercritical, regenerated, closed-loop organic Rankine cycle in a 2 × 2 × 2 ft package.
Four subsystems
Fuel/air. A lift pump draws fuel through 25 ft of line; a metering pump (0.1–0.3 gal/h) and combustion blower share one gearbox so the mixture stays correct; an air-atomising nozzle, magneto igniter and aneroid altitude-compensating valve complete it. At 100 % fire rate the burner releases about 30,000 Btu/h (≈ 8.8 kW) from 2.0 lb/h of fuel, and the heater measured 87 % efficient on Set 1.
Working fluid. CP-25 was selected for thermal stability at 825 °F, a −139 °F freezing point (essential for −65 °F starts) and a moderate condensing pressure that keeps heat exchangers small. Only about 2.4 % of the inventory sits in the hot vaporizer, limiting degradation; the total charge is about 2.4 lb.
Electrical. Two brushless Lundell alternators share the shaft: the main machine generates 2,000 Hz AC rectified to 26–34 V DC for the user; the auxiliary feeds the battery, controls and the constant- and variable-frequency inverters that drive the accessory motors. Both are cooled by the working fluid.
Start and control. Turbine speed is held constant by switching vapour to between 2 and 10 nozzles; burner firing rate follows load through the variable-speed motor. Starting is by battery, external DC or hand crank.
05The turbine and combined rotating unit
Turbine, two alternators and the feed pump on one working-fluid-lubricated shaft inside a hermetic hotwell.
Turbine design data (ATR 1182 data sheet)
| Parameter | Value |
|---|---|
| Type | Single-stage, partial-admission, supersonic axial impulse |
| Inlet | 625 psia (43.1 bar), 825 °F (440 °C) |
| Exhaust | 3.96 psia (0.273 bar) |
| Nozzles | 10 C-D nozzles, 7 active; throat 0.0208 in, area ratio 24.1; 15° angle |
| Jet Mach number | 3.36 (rotor relative inlet 2.04) |
| Wheel | 3.000 in pitch dia., 111 blades, 0.21 in chord, 24°/24° |
| Speed / blade speed | 55,000 rpm / 720 ft/s (219 m/s) |
| U/C₀, Ns, Ds | 0.362, 10.4, 5.33 |
| Efficiency / output | 61 % / 3.43 hp (2.56 kW) shaft |
| Clearances | 0.010 in axial and tip (unshrouded) |
What went wrong with the turbine, and the fix
On Set 1 the nozzles were spaced apart around the arc, so each blade passage saw an intermittent jet. Nitrogen spin-rig tests showed the original plate reached only 48 % of theoretical turbine efficiency; a test plate with nozzles touching edge-to-edge gave a continuous stream and ≥ 84 %. Set 2 received the close-spaced plate, a stiffer hotwell, a larger shaft with bending criticals moved above 55,000 rpm, and the pitot pump moved aft so seal leakage drains to the hotwell instead of quenching the turbine exhaust.
Set 2's turbine then ran at roughly 55–60 % efficiency near design back-pressure (27.6–59.6 % across all runs), close to prediction. The remaining shortfall moved to the pitot pump, whose cheaper Set 2 housing and probe cut its efficiency to 12–17 %.
06The Lundell alternator (NASA TM X-2200)
Why a NASA space-power report belongs with this set.
The ORC set uses Lundell (claw-pole) alternators: a stationary field coil and a solid rotor, so there are no slip rings, brushes or sliding contacts — the only practical choice for a hermetic machine spinning at 55,000 rpm in its own working fluid. NASA's Brayton-cycle space-power work used the same "modified Lundell" or Rice type at 36,000 rpm, 1,200 Hz.
TM X-2200 tested whether such an alternator can start its own turbomachine as an induction motor, with eddy currents in the solid rotor providing the torque. Results for the 10.7 kW machine at 0.1 V/Hz:
| Supply frequency | Synchronous speed | Starting torque | Peak armature current |
|---|---|---|---|
| 400 Hz | 12,000 rpm | ≈ 1.7 p.u. | 3.1 p.u. |
| 800 Hz | 24,000 rpm | ≈ 1.7 p.u. | 4.0 p.u. |
| 1,200 Hz | 36,000 rpm | ≈ 1.5 p.u. | 4.6 p.u. |
1 p.u. torque = 2.09 lb·ft (2.83 N·m). Torque-speed curves resemble a high-resistance-rotor induction motor; 400 Hz gave the most torque per amp, and ≈ 30 V would accelerate the unit to its 12,000 rpm self-sustaining point.
07Test results: Set 1 vs Set 2
Both sets ran, proved the architecture and were delivered to Fort Belvoir; neither reached 1.5 kW net.
| Item | Set 1 (ATR 1182) | Set 2 (AER 1263) | Design |
|---|---|---|---|
| Weight / volume | 193 lb / 7.65 ft³ | 212 lb wet / 7.7 ft³ | ≤ 150 lb / ≤ 8 ft³ |
| Best net output | Not reached; 1.1 kW projected with new nozzles | 477 W (average near zero) | 1,500 W |
| Heater efficiency | 0.88 | 0.82–0.84 | 0.88 |
| Regenerator effectiveness | 0.80 | 0.69–0.74 | 0.855 |
| Pitot pump efficiency | 0.28 | 0.124–0.165 | 0.35 |
| Turbine efficiency | 0.31 | ≈ 0.55–0.60 near design | 0.62 |
| Parasitic (accessory) power | Higher than predicted | 466–605 W (≈ 76 W lower than Set 1) | — |
| Running time | 49 hot tests; 15 h turbo-alternator; 75 h accessories; 2,300+ h bearing rig | 1,500 h CRU bearings; 60 h accessories; 20 h hermetic system | 3,000 h life |
| Start | Accumulator + hand pump | Start pump; automatic start not achieved | < 10 min, automatic |
| Noise | Subsystem data suggest compliance; "quiet" by ear | not yet to spec — gearbox and CRU housing resonance | Inaudible at 100 m |
Sundstrand's improvement path (Table IV-C)
| Cumulative improvement | Net output |
|---|---|
| Turbine at η = 0.58 | 0.35–0.65 kW |
| + reduced heat loss | 0.38–0.68 kW |
| + design regenerator effectiveness | 0.60–0.85 kW |
| + design pitot pump efficiency | 1.1–1.43 kW |
| + design heater efficiency | 1.2–1.57 kW |
| + increased fuel flow | > 1.5 kW |
Projected system thermal efficiency 10.4–13.9 % (HHV). None of these steps was demonstrated within the reported program.
08Analysis: an off-grid home or ranch with a wood-fired burner
Could this design — as built, or as a pattern for a new machine — supply a remote home or ranch, firing on wood instead of diesel or JP fuel?
8.1 Matching the electrical output to the load
The output side is a surprisingly good match. A continuous 1.5 kW is 36 kWh/day, more than a typical efficient off-grid home uses (often 5–15 kWh/day) and roughly a typical grid-connected household. Better still, the 26–34 V DC adjustable output is a natural battery charger for a 24 V battery bank, the standard architecture of off-grid homes: the ORC charges the bank, and an inverter supplies 120/230 V AC. A few hours of running per day at full output would cover a modest home; a ranch with well pumps, refrigeration and workshop loads would need the battery bank to absorb start-up surges.
The catch is what the machine itself consumes. Set 2's accessories drew 466–605 W (blowers, fans, boost pump, inverters, controls), a third of the design output, and the best measured net output was 477 W. At realistic part-load the parasitic share gets worse, not better.
8.2 Wood consumption
Wood-fired heaters are less efficient than the 87–88 % liquid-fuel heater in the reports; a well-designed wood or wood-gasification firebox delivering heat to a high-temperature exchanger is assumed here to reach about 70 %. Scaling the reported system efficiencies by 0.70/0.87 and using air-dried hardwood at ≈ 4.0 kWh/kg (20 % moisture) and ≈ 1,600 kg per cord:
| Case (1.5 kWe net) | Wood-to-electric η | Firing rate | Wood per hour | Wood per day (24 h) | Cords / yr (24 h) | Cords / yr (7 h/day) | Heat to recover |
|---|---|---|---|---|---|---|---|
| Spec minimum (7.5 %) | 6.0 % | 25 kW | 6.2 kg | 149 kg | 34 | 9.9 | ≈ 16 kW |
| Sundstrand projection, low (10.4 %) | 8.4 % | 18 kW | 4.5 kg | 108 kg | 24.5 | 7.2 | ≈ 11 kW |
| Sundstrand projection, high (13.9 %) | 11.2 % | 13 kW | 3.4 kg | 80 kg | 18.4 | 5.4 | ≈ 8 kW |
Assumptions by Infinity Turbine LLC, not from the reports: 70 % wood heater efficiency; 4.0 kWh/kg; 1,600 kg/cord. Heat to recover ≈ heat absorbed by the cycle minus electrical output. For comparison, the as-built Set 2 at 477 W would need roughly three times the wood per kWh.
8.3 Combined heat and power is the only way the numbers work
An ORC at this scale is fundamentally a heater that also makes some electricity: roughly 85–90 % of the wood energy leaves as low-temperature heat. The original set dumps it to air through the condenser at about 100–200 °F, and even provided ducting so the warm exhaust could heat a tent or shelter. On a ranch, replacing the air-cooled condenser with a water-cooled one feeding a hydronic floor, radiators, a domestic hot-water tank or a buffer tank turns that loss into useful heat. In a heating-dominated climate in winter this is attractive; in summer the heat has no use and the electricity alone does not justify the wood.
8.4 Engineering issues in converting the as-built design to wood
| Issue | As-built design | Wood-fired implication | Rating |
|---|---|---|---|
| Heat source control | Atomised liquid fuel; firing rate follows load within seconds via metering-pump speed | A wood fire cannot be turned down or off quickly. Needs a thermal buffer (thermal-oil loop or storage) plus a vapour dump/bypass condenser for load rejection and shutdown. | Major |
| Heater/vaporizer | Compact spiral monotube, high-flux, clean "blue flame" combustion gas | Ash, soot and creosote would foul the compact matrix; hot spots risk cracking the working fluid at 825 °F. Needs a new, larger, cleanable fire-tube or indirect heater. | Major |
| Working fluid | CP-25 at 825 °F supercritical, ~625–975 psia | CP-25 is a 1970s fluid not available today. A modern fluid (e.g. toluene, a siloxane or a hydrocarbon) at lower temperature means a re-designed cycle and turbine. | Major |
| Turbine / CRU | 3 in wheel, 55,000 rpm, 0.010 in clearances, fluid-lubricated bearings | Bespoke, no spares. A rub or bearing failure on a remote ranch means a factory rebuild. Long-life (20,000+ h) operation was never demonstrated. | High |
| Parasitic power | 466–605 W for blowers, fans, pumps, inverters | Wood firing adds draft-fan and feeder loads; a 1.5 kW net machine would need ~2.1–2.3 kW gross. | High |
| Start-up | Battery start; automatic start never achieved | Wood fires take 20–60 min to reach temperature; a hand-tended or automatic pellet/chip start is needed. Fine for a battery-backed system. | Moderate |
| Noise | Designed to be inaudible at 100 m | A real advantage over diesel gensets for homes and livestock; the wood burner itself is quiet. | Benefit |
| Output format | 26–34 V DC, regulated, current-limited | Ideal for charging a 24 V battery bank with an inverter. | Benefit |
| Fuel flexibility | Multi-fuel liquid | Wood, pellets, chips, nut shells or crop residue are all possible with an external combustor — independence from fuel deliveries. | Benefit |
| Safety | Hermetic, flammable organic fluid at high pressure; enclosed burner | Open wood fire next to a hot flammable-fluid loop needs separation (indirect heating loop), pressure relief, over-temperature trips and fire-safe siting. | High |
8.5 Comparison with alternatives for an off-grid ranch
| Option | Strengths | Weaknesses |
|---|---|---|
| Revived 1.5 kW Sundstrand-type ORC on wood | Silent; uses free on-site wood; heat + power; 24 V DC native | Never reached rated output; obsolete fluid; bespoke 55k rpm turbine; heavy fuel handling; complex controls |
| Solar PV + LiFePO₄ battery | Lowest running cost; no fuel; minimal maintenance; modular | Seasonal and weather-dependent; weak in high-latitude winters |
| Diesel / propane genset backup | Cheap, proven, serviceable anywhere | Noisy; fuel deliveries; engine maintenance |
| Purpose-built small wood/biomass ORC or Stirling CHP | Designed for wood from the start; buffer tank and hydronic heat built in | Few products at 1–5 kWe; still costly per kW; electrical efficiency typically single-digit % |
09Verdict
Not as built — but the concept fits a cold-climate, wood-rich ranch as a winter CHP unit beside solar
The Sundstrand set is a remarkable piece of 1970s engineering and the right architecture for quiet, fuel-flexible, battery-charging power. As built, however, it is a development prototype that never delivered more than about a third of its rating, uses an unobtainable working fluid, depends on a fast-responding liquid-fuel burner, and has no proven long-life record. It is not appropriate to install the original hardware, or a straight copy, on an off-grid home or ranch with a wood burner.
A redesigned machine on the same principles is reasonable where three conditions hold together: a long heating season that can use 8–16 kW of rejected heat, a free or cheap on-site wood supply, and a battery bank with solar PV that carries the summer. In that role, running a few hours a day to charge a 24 V bank while heating the house and water, it would use about 5–10 cords a year.
Recommended design changes for a wood-fired revival
- Fire a wood-gasification or pellet boiler into an intermediate thermal-oil (or pressurised-water) loop with a buffer tank, so the ORC sees steady, controllable heat and the fire never touches the organic fluid.
- Drop to a subcritical cycle around 250–300 °C with a currently available, thermally stable fluid; accept a few points of efficiency for life and safety.
- Use a water-cooled condenser feeding hydronic heating and domestic hot water; size the plant to the heat demand and treat electricity as the by-product.
- Keep the 24–28 V DC battery-charging output and add a commercial inverter; start the turbo-alternator by motoring it through its own alternator, as NASA TM X-2200 demonstrated.
- Use the lessons of the program directly: close-spaced (full-arc) nozzles, a stiff rotor with critical speeds above running speed, the feed pump at the aft end, and a lower-speed expander or larger wheel to relax 0.010 in clearances and simplify service.
10References
- McKenna, R. F. Organic Rankine Cycle Silent Power Plant, 1.5 KW, 28 Volts D.C. — Final Report (Set No. 1). Sundstrand Aviation Engineering Report ATR 1182, May 1974. Contract DAAK02-72-C-0472; DTIC/NTIS AD-A000900. Approved for public release. Re-typeset edition 2026 (Infinity Turbine LLC scan).
- McKenna, R. F. Organic Rankine Cycle Silent Power Plant, 1.5 kWe, 28 V DC — Program Report. Sundstrand Aviation Engineering Report AER 1263, 22 September 1975. DTIC AD-A139160. Approved for public release. Re-typeset edition 2026.
- Sundstrand 1.5 kW ORC Silent Power Plant — Axial impulse turbine wheel reconstruction from data sheet ATR 1182. November 2026.
- Repas, D. S.; Frye, R. J. Motor-Starting Characteristics of a Modified Lundell Alternator. NASA TM X-2200, Lewis Research Center, March 1971.
The renders of the power plant and the combined rotating unit on this page are interpretive illustrations prepared for this summary from the published figures; they are not engineering drawings. Off-grid and wood-fuel figures in Section 08 are estimates under the stated assumptions and are not from the source reports.
Download Publications:
MOTOR STARTING CHARACTERISTICS OF A MODIFIED LUNDELL ALTERNATOR Analysis of the Sundstrand ORC Silent Power 1.5 kW System ORC Silent Power 1.5 kW 1182 Publication ORC Silent Power 1.5 kW 1263 PublicationSummary, analysis, and rendered illustrations © 2026 Infinity Turbine LLC · www.infinityturbine.com · greg@infinityturbine.com.
Source reports ATR 1182, AER 1263 and NASA TM X-2200 are U.S. Government-sponsored documents approved for public release; figures reproduced from them are credited in their captions.