Buildings & data centers
Water-cooled centrifugal chiller
Cooling is a heat pump with a known ceiling. Everything under it is waste.
A chiller moves heat from chilled water to condenser water. The temperature gap between them sets the work. Fouled tubes and warm condenser water widen the gap, and the compressor pays.
500 tons, R-134a, 44 °F leaving chilled water
- Full load
- 0.583 kW/ton
- IPLV
- 0.405 kW/ton
- Capacity
- 500 tons
- Second-law efficiency
- 70%
Anatomy and nameplate
- 1Chilled water supply and return temperature. Sets load with the flow reading.
- 2Chilled water flow. Cooling delivered in tons.
- 3Evaporator pressure. Gives the saturated refrigerant temperature.
- 4Condenser pressure. Gives the condensing temperature and the approach.
- 5Condenser water temperatures and flow. Tower side of the lift.
- 6Compressor current. Input power.
- 7Compressor vibration. Bearing and gear condition.
- 8Oil pressure and temperature. Lubrication health.
| Performance | |
|---|---|
| Capacity | 500 tons (1,758 kW cooling) |
| Refrigerant | R-134a |
| Chilled water | 54 → 44 °F at 1,200 gpm |
| Condenser water | 85 → 95 °F at 1,500 gpm |
| Full-load efficiency | 0.583 kW/ton |
| IPLV | 0.405 kW/ton |
| Second-law efficiency | 70% at full load |
| Machine | |
| Compressor | Single-stage centrifugal, variable inlet guide vanes |
| Motor | 450 hp, 4,160 V, 3-phase |
| Evaporator approach | 4 °F |
| Condenser approach | 3 °F clean |
| Tubes | Enhanced copper, 2-pass |
The physics
The equations behind every curve on this page.
Carnot limit
COP = T_evap / (T_cond − T_evap)
The smaller the temperature gap, the less work per ton.
Real machine
COP = η_II · COP_carnot
Second-law efficiency 70% at full load.
kW per ton
kW/ton = 3.517 / COP
0.583 kW/ton at full load, IPLV 0.405.
IPLV
0.01·A + 0.42·B + 0.45·C + 0.12·D
Weighted average over four load points.
Load and lift
Efficiency changes with how hard the chiller is working and how warm the condenser water is.
Efficiency across the load range
Condenser water follows the rating points: 85, 75, 65, 65 °F
- kW per ton at rating conditions
- Full load
- 0.583 kW/ton
- IPLV
- 0.405 kW/ton
A chiller spends most hours at part load, so the part-load points carry most of the weight in the yearly average.
IPLV = 0.01·A + 0.42·B + 0.45·C + 0.12·D
Condenser water temperature sets the lift
- 100% load
- 75% load
- 50% load
The compressor works against the temperature gap between the evaporator and the condenser. A cooler condenser water supply shrinks the gap and the bill.
COP ≈ η_II · T_evap / (T_cond − T_evap)
Fouling
The approach temperature is the cleanest fouling signal. It moves before the efficiency does.
Fouled condenser tubes
75% load, 75 °F entering condenser water
- kW per ton
- Extra power
- 27 kW, about $8,899 a year
Scale on the tubes forces a hotter refrigerant temperature to push the same heat out. The approach temperature shows it before the efficiency does.
approach = T_refrigerant,sat − T_condenser water,leaving
Failure modes
What breaks, what it looks like, what the twin does about it.
| Failure mode | The physics | Signals that show it | How the twin catches it |
|---|---|---|---|
| Condenser tube fouling | Scale adds thermal resistance. Refrigerant condenses hotter. | Condenser approach up. kW/ton up at the same load and tower water. | Expected approach for the current flow and load. The gap is the fouling. |
| Evaporator fouling | Film on the tubes lowers the saturation temperature. | Evaporator approach up. Lower suction pressure. | Same approach check on the cold side. |
| Refrigerant undercharge | Less liquid reaches the expansion valve. | Rising evaporator approach. Capacity loss. Subcooling falls. | Energy balance across both heat exchangers. |
| Compressor bearing and gear wear | Bearing and mesh frequencies are set by geometry. | Vibration lines. Oil temperature. Current. | Tracks band energy at predicted frequencies. |
| Low oil pressure | Filter, pump or foaming problems cut flow to the bearings. | Oil pressure differential and temperature. | Flags a drift before a trip. |
How we catch it early
A physics twin and a PINN, working on one water-cooled centrifugal chiller.
The twin is a set of equations for this machine, tuned to its nameplate. Give it the flow, speed and conditions and it returns what every reading should be. It always gives the same answer for the same inputs.
A physics-informed neural network, or PINN, sits on top. It is trained on the plant’s own data and it is penalized whenever it breaks a physics rule. So it can only learn what the equations leave out: wear, fouling, friction. It cannot invent an answer that the machine could never produce.
Step 1
Measure
Flow, pressure, current, temperature, vibration. Whatever the asset already has.
Step 2
Physics twin
The equations of this machine, set to its nameplate. It says what the readings should be right now.
Step 3
Residual
Measured minus expected. Demand, weather and speed are already taken out.
Step 4
PINN
A small neural network that learns the cause of the gap, under the physics rules.
Step 5
Finding
A named cause, a date, and the kWh or hours at stake. An engineer reviews it.
loss = (measured − predicted)² + λ · (physics residual)²
The second term is the rulebook. It is zero when the network obeys the equations.
Why this beats black-box AI
A black-box model learns patterns from history. It needs examples of each failure, and it goes quiet when the pump runs at a point it has not seen. The twin starts from physics, so it works on day one. The PINN adds what physics cannot know about this particular machine.
That is also what makes the savings countable. The baseline is a calculation, not a guess, so the gap between baseline and meter is a number a customer can check.
What the twin checks on this asset
- kW per ton against the twin at each load
- Condenser and evaporator approach
- Lift and condenser water setpoint
- Energy balance across the evaporator
| Fixed alarm | Black-box AI | Twin + PINN | |
|---|---|---|---|
| Needs failure history to learn | No | Yes, lots | No |
| Says what is wrong | No | Rarely | Yes, by name |
| Works at a flow it has never seen | No | Often fails | Yes, the equations still hold |
| Puts a number on the saving | No | No | Yes, in kWh and dollars |
| Lead time on slow wear | Late | Varies | Weeks to months |
| Every number can be traced | Yes | No | Yes, to an equation |
What the twin puts a number on
Energy and cost, from the equations.
Full-load efficiency
0.583 kW/ton
500 tons, 85 °F condenser water
IPLV
0.405 kW/ton
Four-point weighted average
Fouled approach, 10 °F
27 kW
$8,899 a year at 3,000 hours
What we do for you
We watch it. We catch it. We prove it.
Our reliability engineers run the twin on your water-cooled centrifugal chiller, review every finding, and help your crew fix it. After the repair we compare the meter to the baseline and write the result in the ledger.