playbooks

home / skills / a5c-ai / babysitter / kinetic-modeler

kinetic-modeler skill

/plugins/babysitter/skills/babysit/process/specializations/domains/science/chemical-engineering/skills/kinetic-modeler

This skill develops and validates reaction kinetics models, estimates parameters from data, and supports reactor design with rigorous uncertainty analysis.

npx playbooks add skill a5c-ai/babysitter --skill kinetic-modeler

Review the files below or copy the command above to add this skill to your agents.

Files (1)
SKILL.md
1.8 KB
---
name: kinetic-modeler
description: Reaction kinetics modeling skill for parameter estimation, mechanism validation, and rate equation development
allowed-tools:
  - Read
  - Write
  - Glob
  - Grep
  - Edit
  - Bash
metadata:
  specialization: chemical-engineering
  domain: science
  category: Reaction Engineering
  skill-id: CE-SK-006
---

# Kinetic Modeler Skill

## Purpose

The Kinetic Modeler Skill develops and validates reaction kinetics models, performing parameter estimation from experimental data and supporting reactor design.

## Capabilities

- Rate equation formulation (power law, LHHW, Eley-Rideal)
- Parameter estimation via nonlinear regression
- Arrhenius parameter calculation
- Activation energy determination
- Model discrimination (AIC, BIC criteria)
- Confidence interval estimation
- Reaction mechanism validation
- Kinetic data analysis

## Usage Guidelines

### When to Use
- Developing kinetic models
- Estimating rate parameters
- Validating reaction mechanisms
- Supporting reactor design

### Prerequisites
- Experimental data available
- Proposed mechanism identified
- Operating conditions characterized
- Thermodynamic constraints known

### Best Practices
- Use statistically valid data
- Test multiple model forms
- Validate with independent data
- Report parameter uncertainties

## Process Integration

This skill integrates with:
- Kinetic Model Development
- Reactor Design and Selection
- Catalyst Evaluation and Optimization

## Configuration

```yaml
kinetic-modeler:
  model-types:
    - power-law
    - langmuir-hinshelwood
    - eley-rideal
    - mechanistic
  estimation-methods:
    - least-squares
    - maximum-likelihood
    - bayesian
```

## Output Artifacts

- Kinetic models
- Parameter estimates
- Confidence intervals
- Model validation reports
- Mechanism analysis

Overview

This skill builds and validates chemical reaction kinetics models to support parameter estimation, mechanism testing, and reactor design. It produces rate equations, fits parameters to experimental data, and quantifies uncertainty for informed engineering decisions. The outputs include validated kinetic models, parameter estimates, and model comparison metrics.

How this skill works

The skill formulates candidate rate laws (power-law, Langmuir–Hinshelwood, Eley–Rideal, mechanistic) and fits them to supplied experimental data using nonlinear regression or Bayesian approaches. It computes Arrhenius parameters, activation energies, confidence intervals, and model selection metrics (AIC, BIC). It also tests proposed mechanisms against data and reports validation results for integration into reactor design workflows.

When to use it

  • Developing kinetic expressions from batch or flow experimental data
  • Estimating reaction rate constants and Arrhenius parameters
  • Comparing competing mechanisms or rate-law forms
  • Validating a proposed mechanism before scaling or reactor design
  • Quantifying parameter uncertainty for sensitivity or optimization studies

Best practices

  • Provide high-quality, statistically valid experimental data with known operating conditions
  • Test multiple model structures and use information criteria (AIC/BIC) for model discrimination
  • Reserve independent data for validation rather than using the same set for fitting
  • Report parameter uncertainties and confidence intervals alongside point estimates
  • Include thermodynamic constraints and physically meaningful bounds on parameters

Example use cases

  • Fit power-law and LHHW models to catalytic reaction data and select the best model by AIC
  • Estimate pre-exponential factors and activation energy from temperature series using Arrhenius fits
  • Validate a proposed elementary-step mechanism by testing predicted rates against experimental profiles
  • Produce confidence intervals for rate constants to support robust reactor design
  • Compare least-squares and Bayesian parameter estimates to assess uncertainty under sparse data

FAQ

What input data do I need?

Time-resolved concentration measurements or steady-state rates, temperatures, pressures, and experimental conditions plus an initial proposed mechanism if available.

Which estimation methods are supported?

Nonlinear least-squares, maximum-likelihood, and Bayesian estimation methods are supported for parameter fitting and uncertainty quantification.

How does the skill choose between competing models?

It calculates information criteria (AIC, BIC) and examines residuals and confidence intervals to recommend the most parsimonious model that fits the data.