# GX230 Technology

The GX230 is a 200 kW modular multi-fuel power unit. Every piece is engineered to work together.

## Key specifications

| Parameter | Value |
|---|---|
| Electrical output | 200 kW |
| Mechanical input | 230 kW |
| Engine thermal efficiency | 42% (BSFC 185 g/kWh at 2,200–3,200 RPM) |
| Engine-to-grid efficiency | 87% |
| Total fuel-to-electrical efficiency | 36.5% |
| Dimensions | 1.4 m × 1.0 m × 2.0 m (~15 ft²) |
| Weight | 700 kg |
| Power density | ~90 kVA/m³ (roughly 5× conventional gensets) |
| Noise level | 69 dB at 5 m |
| Ramp-up time | < 5 seconds |
| Deployment time | ~4 weeks typical |

## Fuels supported

Natural gas, CNG, LPG, synthetic fuel, biofuel, hydrogen blends. Fuel-agnostic by design — customers switch fuels based on local availability and price without changing the unit.

## Operating modes

Standalone, grid-tied, microgrid. A grid-forming inverter makes islanded operation native rather than an afterthought.

## Control system

Bosch ECU plus the proprietary Immedia Power OS. Power OS handles load-following, fuel switching, grid synchronization, and remote diagnostics.

## Economics — illustrative

- Natural gas at $4/MMBTU → end-use electricity at ~$0.037/kWh
- Maintenance: ~$10,000/year per unit at 8 hrs/day operation
- Full unit economics modeled on your site parameters — use the [Fleet Charging Simulator](/simulator/) or [contact sales](/contact/) for a detailed proposal.

## Platform: Immedia Power Operating System

The GX230 is powerful as a single 200 kW unit. A fleet of GX230s across dozens of customer sites becomes a managed, anticipatory infrastructure asset. The Immedia Power platform observes every sensor, predicts every failure, and shapes every dispatch decision.

### Predictive Maintenance: Uptime Assurance

Each unit streams its full sensor signature—vibration, temperature, pressure, electrical harmonics—to the platform in real time. The system establishes a factory baseline for each individual machine, then tracks every deviation. Bearing wear, fuel delivery drift, thermal stress: we detect them weeks before alarm thresholds trip.

At one customer site running an 18-unit fleet, the platform identified a bearing drift at hour 3,840 of operation—280 hours before the unit would have failed. Maintenance was scheduled during a planned downtime window, not as an emergency. Key metrics: 18 active fleet units, 3 pending advisories, 92% model confidence.

### Demand Analysis & Load Forecasting

Grid demand follows patterns: daily cycles, day-of-week signatures, seasonal swings, and the customer's own operational calendar. Per-site load forecasts use weather data, historical consumption, and the facility's own operational schedule.

A 72-hour forecast that refreshes every 15 minutes lets dispatch decide how many units to bring online, how hard to run each one, and whether to bank fuel or hold capacity. Forecast accuracy across five active customer sites: 94.2% over a rolling 30-day window. Key metrics: 5 sites with active forecasts, 72-hour horizon, 15-minute refresh cadence.

### Peak Shaving: Grid Independence

Demand charges—fees the utility imposes for any 15-minute interval in which the site draws above a threshold—are often the largest component of a facility's electricity bill. The platform monitors real-time site load and automatically dispatches GX230 capacity to keep grid import below a pre-set cap.

From the grid operator's perspective, the load curve flattens. From the facility operator's perspective, the load is unchanged—the GX230s absorb the peaks behind the meter. One commercial customer capped grid import at 40% of total site load, reducing demand charges by 58% while adding zero load on the facility floor. Key metrics: 47 peak-shaving events this week, 312 kW offset now, grid draw reduced 58%.

### Islanding & Grid Independence

A GX230 with a grid-forming inverter can run the site as a standalone microgrid. If the grid drops, the island mode activates in under 50 milliseconds. The site load transitions from grid supply to GX230 supply before the utility voltage has even settled to zero.

Multiple units can be paralleled into the same microgrid. A four-unit site operating in island mode sees redundancy: if one unit trips, the other three ramp automatically to fill the gap. This matters for data centers, EV charging hubs, and any facility where a grid outage means lost revenue. Key metrics: Grid-forming inverter topology, 4 units paralleled, transfer time 38 ms.

### Composite Site Score: One Number for Operations

Operators and executives ask the same question differently:
- **Operators:** "Where should I look first today?"
- **Executives:** "How is this site performing?"

The platform answers both with a single rolling composite score—0 to 100—built from five equal pillars: availability (uptime percentage), efficiency (kWh out per unit of fuel in), emissions (grams CO2 per kWh), maintenance posture (scheduled vs. reactive repairs), and orchestration quality (forecast accuracy, demand-charge avoidance, transfer reliability).

One facility's score sits at 94 out of 100. The fleet average is 87. The top quartile score 91 or higher. Within each pillar, the platform surfaces which specific axis is drifting, so operators can drill down to specifics. Key metrics: 5 sites ranked, fleet average 87, top quartile 91+.

## See also

- [Fleet Charging Simulator](/simulator/)
- [Market applications](/market/)
- [Contact sales](/contact/)