Hyprisma Join the waitlist
Menu

How it works

One project, from a first screening toward a bankable study.

A project moves through stages freely and keeps its history. Each stage adds one layer to the one before it, and every number on this page is the app’s own example, solved by its engines.

One plant, four engines · the cost of a tonne of ammonia

One design, solved four ways: the example’s feasibility design v2, 250 MW wind, 200 MWp solar, 20 MW / 80 MWh battery, 150 MW electrolyser, 50 t hydrogen store, 150,000 t/yr ammonia unit. Each rung adds one layer to the one above it, so the gap between two rungs is that layer, not a different plant.

  1. Prefeasibility Annual closed form: the year from its averages

    763 $/t 112,742 t/yr · curtailed 15.0%

    binding: renewables

  2. adds the hours of a real year

    Prefeasibility Hour by hour simple priority rules over a year of weather

    885 $/t 97,494 t/yr · curtailed 26.2%

    binding upstream: renewables

  3. adds operating limits

    Feasibility Rule-based operating limits, no foresight: the base case

    989 $/t 87,015 t/yr · curtailed 34.1%

    Renewables-limited most days · 96 h below its minimum run for lack of hydrogen

  4. adds perfect foresight

    Feasibility LP solver operating limits, the whole year known: the ceiling

    879 $/t 97,866 t/yr · curtailed 25.8%

    never below its minimum run

  5. adds every unit in detail, and compliance

    Pre-FEEDComingnot built yet

  6. adds cash flows, debt and returns

    FinancingComingnot built yet

The same plant, four ways: adding the hours of a real year takes the cost of a tonne from 763 to 885 $/t; adding operating limits takes it to 989 $/t; adding perfect foresight brings it back to 879 $/t.

Computed by Hyprisma 1.0.5 · example project ‘Ammonia export plant’ · feasibility design v2 · default assumption book v2 · public typical-year dataset (2024) · computed 12 November 2026

What each stage adds

  1. 01Available

    Prefeasibility

    Adds time. A first screening of the plant, in seconds.

    Annual
    The year from its averages, in closed form. Names the stage that binds among three limits.
    Hour by hour
    The same plant over a year of real weather, run by simple priority rules: every unit free to run from zero to its size.
    You read
    The cost of each product per tonne, what limits output, the energy that goes unused, the right size of each stage.
  2. 02Available

    Feasibility

    Adds operating limits and the best dispatch. The plant as it has to run.

    Operating limits
    Minimum load, ramps and standby for the electrolyser and each unit; the battery and the hydrogen store as they charge and discharge.
    Two engines
    The rule-based engine runs the plant as an operator would, with no foresight: the base case. The LP solver, an optimiser that solves a linear program, knows the whole year: the ceiling.
    You read
    Hours short of hydrogen and of power, kept apart; storage cycles; the year hour by hour; the cost of each product.
  3. 03Coming

    Pre-FEED

    Adds plant fidelity. Every unit of the plant in detail, and compliance.

  4. 04Coming

    Financing

    Adds the financial model. Cash flows, debt and returns, as a lender reads them.

The three limits

A plant makes no more than its weakest stage allows. A screening computes three ceilings, the energy, the electrolyser and the product unit, and names the lowest: the stage that limits the plant, the binding one.

Three limits, one answer

The example’s prefeasibility design v1, annual resolution: the closed form names the stage that limits the plant, the binding one.

Electrolyser 150 MW, the example’s design

  1. Wind and sun 250 MW wind and 200 MWp solar: the ammonia their energy can make

    112,742 t/yr binding

  2. Electrolyser its size at 80.0% maximum use: the ammonia its hydrogen can make

    115,570 t/yr binding

  3. Ammonia unit 150,000 t/yr at 85.0% maximum use

    127,500 t/yr binding

The plant makes the lowest: 112,742 t/yr of ammonia, at 736 $/t.

At 150 MW the wind and sun set the lowest ceiling, 112,742 t/yr, so the plant makes 112,742 t of ammonia a year, at 736 $/t; on their own, the electrolyser 115,570 t/yr and the ammonia unit 127,500 t/yr.

Computed by Hyprisma 1.0.5 · example project ‘Ammonia export plant’ · prefeasibility design v1 · annual resolution · default assumption book v2 · public typical-year dataset (2024) · computed 12 November 2026

Two engines bracket the truth

In feasibility the same plant runs twice: by rules that never look ahead, and by an optimiser that sees the whole year. A real plant, run well, lands between the two.

One week, hour by hour · 1–7 April 2024

The example’s feasibility design: 250 MW wind, 200 MWp solar, 20 MW / 80 MWh battery, 150 MW electrolyser, 50 t hydrogen store. The week is the full Monday-to-Sunday week whose wind and solar energy is closest to the median week of the year.

  • Wind available
  • Solar available
  • Electrolyser draw, its 1.5 MW standby included
  • Electrolyser size, 150 MW
  • Battery
  • Hydrogen store

Point at the chart, or focus it and use the arrow keys, to read any hour.

Rule-based engine no foresight · the base case

storage level, % of usable capacity

LP solver perfect foresight · the ceiling

storage level, % of usable capacity

In this week the LP solver makes 10.3% more hydrogen than the rule-based engine (344.7 t against 312.5 t): it empties and refills the battery on 7 of 7 days, where the rule-based engine never lets it fall below 63%.

The whole typical year, the same design · cost of a tonne of ammonia

  • Rule-based engine the base case 989 $/t
    87,015 t/yr
    96 h short of hydrogen
  • LP solver the ceiling 879 $/t
    97,866 t/yr
    0 h short of hydrogen

Over the year the same plant makes 87,015 t of ammonia at 989 $/t run by the rules, 96 h below its minimum run for lack of hydrogen, and 97,866 t at 879 $/t run by the solver, never below its minimum run: the first is the base case, the second the ceiling.

Every product, one upstream plant

One upstream plant, every product

A project makes one product or several from the same electrolyser; in prefeasibility each product takes hydrogen in proportion to its size. The CO₂ comes from point-source capture, direct air capture or a bought-in supply.

Shared upstream

  1. Wind
  2. Solar
  3. Battery
  4. Electrolyser
  5. Desalination
  • Ammonia
    1. Air separation (nitrogen)
    2. Haber–Bosch synthesis
  • e-Methanol
    1. CO₂ source
    2. Methanol synthesis
  • e-SAF, Fischer–Tropsch
    1. CO₂ source
    2. Purification and storage
    3. Reverse water-gas shift
    4. Fischer–Tropsch
    5. Upgrading
  • e-SAF, methanol-to-jet
    1. CO₂ source
    2. Purification and storage
    3. Methanol synthesis
    4. Methanol-to-jet
  • e-NG
    1. CO₂ source
    2. Methanation
  • Green iron
    1. Iron ore
    2. Direct reduction with hydrogen
  • Green steel
    1. Iron ore
    2. Direct reduction with hydrogen
    3. Electric arc furnace
  • Exported hydrogen
    1. Injection compressor