Skip to content

Umgebung in aktiver Entwicklung: Es können Änderungen oder unvollständige Funktionen auftreten.

JAOT

Chemical Process Optimization

Chemical process optimization helps engineers formulate blends and maximize reactor yields under strict purity and safety specifications. The example below shows feedstock blending: given multiple raw materials with different costs and quality properties, find the least expensive combination that meets product specifications.

Step-by-Step Walkthrough

1. Define feedstocks and their properties

List each raw material or intermediate chemical with its cost, availability, and relevant quality properties (purity, density, viscosity, sulfur content, etc.).

2. Specify product quality requirements

Define the target product specifications with minimum and maximum bounds. For example, a gasoline blend must have an octane rating of at least 87 and a sulfur content below 30 ppm.

3. Model process constraints

Include physical and operational limits:

  • Reactor temperature and pressure ranges
  • Throughput capacity limits
  • Storage tank capacities
  • Environmental emission limits
  • Safety margins on hazardous material concentrations

4. Set the optimization objective

Common objectives include:

  • Minimize feedstock cost per unit of product
  • Maximize yield from a given set of inputs
  • Minimize energy consumption while maintaining throughput

Chemical Process in JAOT Builder

5. Validate and implement

Review the optimal blend ratios or operating parameters. Verify they respect all safety constraints. Run the solution against historical data to confirm expected improvements before implementing on the plant floor.

Example Parameters

import httpx

API_URL = "https://jaot.io/api/v2"
headers = {"Authorization": "Bearer ok_live_your_key_here"}

# Minimize cost of blending 4 feedstocks to meet product specs
response = httpx.post(f"{API_URL}/solve", headers=headers, json={
    "variables": [
        {"name": "feedstock_a_liters", "type": "continuous", "lb": 0, "ub": 5000},
        {"name": "feedstock_b_liters", "type": "continuous", "lb": 0, "ub": 3000},
        {"name": "feedstock_c_liters", "type": "continuous", "lb": 0, "ub": 4000},
        {"name": "feedstock_d_liters", "type": "continuous", "lb": 0, "ub": 2000},
    ],
    "objective": {
        "sense": "minimize",
        "coefficients": {
            "feedstock_a_liters": 0.85,
            "feedstock_b_liters": 1.20,
            "feedstock_c_liters": 0.95,
            "feedstock_d_liters": 1.50,
        },
    },
    "constraints": [
        {
            "name": "total_batch_volume",
            "coefficients": {
                "feedstock_a_liters": 1, "feedstock_b_liters": 1,
                "feedstock_c_liters": 1, "feedstock_d_liters": 1
            },
            "sense": "==",
            "rhs": 10000,
        },
        {
            "name": "min_purity",
            "coefficients": {
                "feedstock_a_liters": 0.92, "feedstock_b_liters": 0.98,
                "feedstock_c_liters": 0.88, "feedstock_d_liters": 0.99
            },
            "sense": ">=",
            "rhs": 9300,
        },
        {
            "name": "max_sulfur_content",
            "coefficients": {
                "feedstock_a_liters": 0.005, "feedstock_b_liters": 0.001,
                "feedstock_c_liters": 0.008, "feedstock_d_liters": 0.0005
            },
            "sense": "<=",
            "rhs": 30,
        },
        {
            "name": "min_viscosity",
            "coefficients": {
                "feedstock_a_liters": 4.2, "feedstock_b_liters": 6.1,
                "feedstock_c_liters": 3.8, "feedstock_d_liters": 7.5
            },
            "sense": ">=",
            "rhs": 45000,
        },
    ],
})
result = response.json()

print(f"Minimum blend cost: ${result['objective_value']:.2f}")
for var in result["variables"]:
    if var["value"] > 0:
        print(f"  {var['name']}: {var['value']:.0f} liters")

Templates

  • Custom Optimization -- build a blending or process optimization model tailored to your chemical engineering requirements

Next Steps