Predictika · FuelCell Energy Block Systems

An AI sales agent's fuel cell system order, checked before it goes out

An AI sales agent prepares an order for FuelCell Energy Block Systems from public data sheets: how many systems, which fuel, the grid connection and the site. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of those data sheets and returns a verdict, every reason and the figures it computed.

Scope: FuelCell Energy 1.25 MW and 2.5 MW FuelCell Energy Block Systems: fuel, hydrogen and biogas limits, grid connection, site and total output. See everything this validator covers

This validator models orders for the two FuelCell Energy Block Systems in FuelCell Energy's April 2026 data sheets (1.25 MW and 2.5 MW), not just the scenarios shown below.

  • 1.25 MW Energy Block System — 1.25 MW net at 480 VAC, 60 Hz; natural gas, biogas (55 to 100 percent methane, pre-conditioned, rated at 60 percent methane) or natural gas and hydrogen blends (0 to 50 percent hydrogen); fuel inlet 15 to 20 psig; listed footprint 58 x 42 ft.
  • 2.5 MW Energy Block System — 2.5 MW net at 13.8 kVAC, 60 Hz; the same fuel options and limits; fuel inlet 15 to 20 psig; listed footprint 72 x 60 ft.
  • Both — outdoor operation from -20 to 120 degrees F; other medium voltages are listed as available on request, so an order for one needs approval.
  • Figures checked — the total output the order states is compared with the number of systems times the listed output of each system.

Any order for these two systems that touches these rules is in scope — not only the scenarios pictured here. Not covered: prices (FuelCell Energy publishes none for these products, so this validator produces no price build-up or line-item quote), the 12.5 MW block and other FuelCell Energy products, carbon capture, service and power purchase agreements, and any change FuelCell Energy has made to its products since these sheets were published.

The AI sales agentBuilds the fuel cell system order from the conversation and says it is ready.
Predictika validatorChecks every rule at once and returns the verdict, each finding, and the computed figures.

7 scenarios in 3 groups · each group runs from the agent’s first attempt to the outcome · click one to switch

What the agent proposes

The AI sales agent says

Illustrative agent statement. The validator receives the structured request below.

Proposing agent: AI sales agent (scripted)

Before sending, the agent must call Recorded run

Predictika validator


      
Waiting for the call.

Where rules-based CPQ gets it wrong

Click to expand

Our assessment of what a typical rules-based CPQ (product rules, price rules and validation rules, as in Salesforce CPQ) does with the same cases. It is a judgment about effort and reliability, not a claim that nothing can be built.

What the validator caughtTypical rules-based CPQPredictika
Biogas ordered without the pre-conditioning the sheet requiresOnly with care
Fuel type and the pre-conditioning requirement sit in different places, so a rule has to link them for each plant model.
The fuel and the pre-conditioning are in one request and compared with the sheet on every call.
Hydrogen share above the listed 50 percent, and a methane share that differs from the 60 percent rating basisOnly with care
The hydrogen limit applies only to blends and the methane rating basis only to biogas, so each needs its own fuel-specific rule and message.
Both are checked on every call: a block above 50 percent hydrogen, an approval when biogas methane differs from the rating basis.
Grid voltage that does not match the system (480 V against 13,800 V)Only with care
A per-model voltage table has to be maintained and compared with a free-entry voltage.
The interconnection voltage is compared with the system on every call: a block for low voltage, an approval for another medium voltage.
Total output stated by the agent that does not equal systems times outputOnly with workarounds
The stated figure is usually free text or a separate field, so it takes an extra rule that multiplies and compares.
The stated figure is part of the request and is compared with the computed figure on every call.
Every problem returned in one answerOnly with workarounds
Rules tend to fire as a step is completed, so a rep often fixes one error and meets the next.
All findings come back together, whatever order the choices arrived in.

What this validator catches: findings from its test cases

Click to expand

Every rule is checked on every call, in any order the choices arrive; one problem never hides another. Blocking findings stop the quote; approval findings need a person to sign off; everything else is computed.

1The system outputs 480 V, 60 Hz as listed; the order specifies 13,800 V, a medium voltage the data sheet offers only on request.
2The system outputs 13,800 V, 60 Hz as listed, and the data sheet offers only other medium voltages on request; the order specifies a 480 V interconnection.
3The data sheets list 60 Hz; the order specifies 50 Hz.
4The order has 0 units; at least one is needed.
5The systems are listed for outdoor operation from -20 to 120 degrees F; the site ranges from -25 to 100 degrees F.
6The systems are listed for outdoor operation from -20 to 120 degrees F; the site ranges from -10 to 125 degrees F.
7The pad (50 x 60 ft) is smaller than the listed footprint of one system (58 x 42 ft).
8The pad (75 x 50 ft) is smaller than the listed footprint of one system (72 x 60 ft).
9The pad (58 x 42 ft) is smaller than the listed footprint of one system (72 x 60 ft).
10The systems are listed for 0 to 50 percent hydrogen blending; the order has 55 percent.
11A hydrogen share of 20 percent is only listed for natural gas and hydrogen blends.
12A hydrogen share of 10 percent is only listed for natural gas and hydrogen blends.
13Biogas must be pre-conditioned to meet the fuel composition requirements of the system; the order has no pre-conditioning.
14The systems list biogas with 55 to 100 percent methane; the order has 50 percent.
15The sheet ratings for biogas are based on 60 percent methane; the order has 55 percent, so delivered performance may differ.
16The sheet ratings for biogas are based on 60 percent methane; the order has 100 percent, so delivered performance may differ.
17The sheet ratings for biogas are based on 60 percent methane; the order has 70 percent, so delivered performance may differ.
18The systems list a fuel inlet pressure of 15 to 20 psig; the order has 12 psig.
19The systems list a fuel inlet pressure of 15 to 20 psig; the order has 25 psig.
20The order states 4,000 kW in total, but 4 system(s) at 1,250 kW each give 5,000 kW.
21The order states 7,000 kW in total, but 3 system(s) at 2,500 kW each give 7,500 kW.
22A hydrogen share of 45 percent is only listed for natural gas and hydrogen blends.

Scenarios and results

Click to expand

The recorded result of the validator for each scenario on this page: the request the AI agent sent, the verdict, and every finding, in the order they appear above.

1. Biogas systems: Four 1.25 MW systems on biogas

The agent's request. Please quote four 1.25 MW FuelCell Energy Block Systems running on biogas, connected at 480 V, for 5,000 kW in total. The pad is 62 x 46 ft. It is ready to send.

Verdict: Blocked.

  • Biogas not pre-conditioned. Biogas must be pre-conditioned to meet the fuel composition requirements of the system; the order has no pre-conditioning.

2. Biogas systems: The agent resubmits after fixing the errors

The agent's request. Please quote four 1.25 MW FuelCell Energy Block Systems running on biogas, connected at 480 V, for 5,000 kW in total. The pad is 62 x 46 ft. The biogas is pre-conditioned. It is ready to send.

Verdict: Allowed.

  • Order summary. 4 system(s) at 1,250 kW each, 5,000 kW in total.

3. Hydrogen blend: Two 2.5 MW systems on a hydrogen blend

The agent's request. Please quote two 2.5 MW FuelCell Energy Block Systems running on a natural gas and hydrogen blend with 55 percent hydrogen, connected at 480 V, for 5,000 kW in total. The pad is 75 x 65 ft. It is ready to send.

Verdict: Blocked.

  • Grid voltage not offered for this system. The system outputs 13,800 V, 60 Hz as listed, and the data sheet offers only other medium voltages on request; the order specifies a 480 V interconnection.
  • Hydrogen share above the listed range. The systems are listed for 0 to 50 percent hydrogen blending; the order has 55 percent.

4. Hydrogen blend: The agent resubmits after fixing the errors

The agent's request. Please quote two 2.5 MW FuelCell Energy Block Systems running on a natural gas and hydrogen blend with 55 percent hydrogen, connected at 13,800 V, for 5,000 kW in total. The pad is 75 x 65 ft. It is ready to send.

Verdict: Blocked.

  • Hydrogen share above the listed range. The systems are listed for 0 to 50 percent hydrogen blending; the order has 55 percent.

5. Hydrogen blend: The agent resubmits after fixing the errors

The agent's request. Please quote two 2.5 MW FuelCell Energy Block Systems running on a natural gas and hydrogen blend with 45 percent hydrogen, connected at 13,800 V, for 5,000 kW in total. The pad is 75 x 65 ft. It is ready to send.

Verdict: Allowed.

  • Order summary. 2 system(s) at 2,500 kW each, 5,000 kW in total.

6. Microgrid: Three 1.25 MW systems for a microgrid

The agent's request. Please quote three 1.25 MW FuelCell Energy Block Systems running on natural gas, connected at 480 V, for 4,000 kW in total. The pad is 62 x 46 ft. It is ready to send.

Verdict: Blocked.

  • Stated total output differs. The order states 4,000 kW in total, but 3 system(s) at 1,250 kW each give 3,750 kW.

7. Microgrid: The agent resubmits after fixing the errors

The agent's request. Please quote three 1.25 MW FuelCell Energy Block Systems running on natural gas, connected at 480 V, for 3,750 kW in total. The pad is 62 x 46 ft. It is ready to send.

Verdict: Allowed.

  • Order summary. 3 system(s) at 1,250 kW each, 3,750 kW in total.

Narration transcript

Click to expand

The text of the narrated walkthrough, step by step.

  1. Welcome to a demo of Predictika's Validator. An AI sales agent prepares an order for FuelCell Energy Block Systems from public data sheets: how many systems, which fuel, the grid connection and the site. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of those data sheets and returns a verdict, every reason and the figures it computed.
  2. I will start with the first request by the agent.
  3. The agent says. Please quote two 2.5 MW FuelCell Energy Block Systems running on a natural gas and hydrogen blend with 55 percent hydrogen, connected at 480 V, for 5,000 kW in total. The pad is 75 x 65 ft. The rest of the request is on screen.
  4. Before it goes out, the agent calls the validator.
  5. Blocked, with two findings, all in one answer. The system outputs 13,800 V, 60 Hz as listed, and the data sheet offers only other medium voltages on request; the order specifies a 480 V interconnection. The systems are listed for 0 to 50 percent hydrogen blending; the order has 55 percent.
  6. The validator returned two blocking findings in one answer. A check that reads one field at a time would pass most of this; it shows only when the choices in the quote are checked against each other.
  7. What this shows: for hydrogen share above the listed 50 percent, and a methane share that differs from the 60 percent rating basis, a rules-based CPQ is error-prone for this.
  8. Now the agent resubmits.
  9. The corrected request. It changes the grid connection to 13,800 V.
  10. It calls the validator again.
  11. Still blocked, with one finding, all in one answer. The systems are listed for 0 to 50 percent hydrogen blending; the order has 55 percent.
  12. One finding cleared. Still blocking: hydrogen share above the listed range.
  13. The agent resubmits once more.
  14. The revised request. It changes the hydrogen share to 45 percent.
  15. It calls the validator once more.
  16. This time the validator clears it.
  17. The earlier finding is cleared, and the validator's figure matches the quote.
  18. Two problems caught in one call, fixed in two passes, and a clean quote. That is the value of Predictika's validator: catching mistakes before they reach the customer. Now you can try the other scenarios yourself.