skills.rest
equinor
equinorOFFICIAL
Explore all skills by this author →
WebsiteGitHub

neqsim-benchmark-reference-data

Validates NeqSim model results against independent reference data with graded PASS/WARN/FAIL comparisons.

4|2|Updated May 31, 2026
Search Tags:#reference data#thermodynamics#model verification#neqsim#coolprop#benchmark validation#equation of state
One-click install
npx skills add https://github.com/equinor/neqsim-community-skills --skill neqsim-benchmark-reference-data-equinor
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: neqsim-benchmark-reference-data
Source: https://github.com/equinor/neqsim-community-skills/tree/main/skills/pvt/benchmark-reference-data
Command: npx skills add https://github.com/equinor/neqsim-community-skills --skill neqsim-benchmark-reference-data-equinor

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires CoolProp.

What problem does it solve? Benchmark validation in NeqSim tasks is usually rewritten from scratch in every notebook, with reference values pasted as bare literals, no citation, invented tolerances, and no check that the reference is actually independent of the model being tested. This Skill supplies the reference side of the comparison: traceable sources, applicability ranges, grading rules, and the exact benchmark_validation block that report generators and CI gates consume. ## Core Features & Use Cases - Reference source registry: Registers IAPWS-95, IAPWS-IF97, Span-Wagner CO2, Setzmann-Wagner methane, Span nitrogen, Bucker-Wagner ethane, Lemmon propane, GERG-2008, CoolProp HEOS, and NIST WebBook with authority tiers, validated temperature/pressure ranges, and stated uncertainties. - Offline anchor table: Published critical points, triple points, boiling points, and an ambient water density that run with no optional dependency and no network. - Optional CoolProp backend: Generates reference values at arbitrary (T, P) states or grids when CoolProp is installed. - Comparison and reporting layer: Grades PASS/WARN/FAIL/INFO, rejects references that do not outrank the model basis, checks deviations against source uncertainty, enforces a three-point minimum, and emits the benchmark_validation block for results.json plus a Markdown table. - Use Case: While writing a benchmark notebook for an SRK pure-component study, compare NeqSim critical constants against the anchor table, print the Markdown table, and write the results.json block that passes the CI gate. ## Quick Start Ask the agent to benchmark my NeqSim CO2 critical temperature and pressure against the offline anchor reference data and produce the benchmark_validation block for results.json.

Frequently Asked Questions about neqsim-benchmark-reference-data

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I validate NeqSim results against reference data in Python?▼

Use the compare function with a model value and a ReferencePoint from find_anchor or the CoolProp backend, then add results to a BenchmarkReport. Call to_results_json() to emit the benchmark_validation block and to_markdown() for the report table.

What reference data sources are available for benchmarking thermodynamic models?▼

The registry includes IAPWS-95 and IAPWS-IF97 for water, Span-Wagner for CO2, Setzmann-Wagner for methane, Span for nitrogen, Bucker-Wagner for ethane, Lemmon for propane, GERG-2008 for natural gas mixtures, CoolProp HEOS, and the NIST WebBook, each with validated ranges and stated uncertainties.

Does the benchmark validation work without CoolProp installed?▼

Yes, the offline anchor table works with no optional dependency and no network. It covers critical points, triple points, boiling points, and one ambient water density; CoolProp is only needed for reference values at arbitrary states.

Why is my benchmark comparison graded INFO instead of PASS?▼

A comparison is graded INFO when the reference does not outrank the model basis, when it is marked informational, or when the reference value is zero. Comparing GERG-2008 against a CoolProp HEOS evaluation of the same formulation is a consistency check, not a benchmark.

What are the limitations of the offline anchor table?▼

The anchor table is intentionally small and pure-component only, serving as a smoke test rather than a validation dataset. It contains no mixture data, and a real study should extend it with CoolProp, lab data, or a published case for the fluid actually modelled.

Related Skills

Explore More Software Engineering Skills→
equinorequinor
138

neqsim-api-patterns

Provide reusable NeqSim API patterns for Java and Python thermodynamic calculations.

Official
Intermediate
microsoftmicrosoft
35

quality-gates

Validate and enforce quality gates across CI/CD workflows with pass/fail reports.

Official
Advanced
AndreaMentastiAndreaMentasti
1

qa-quarto

Compare Quarto HTML slides against Beamer PDF benchmarks for content parity.

Community
Advanced
SoljournerSoljourner
45

thermo-package

Compute temperature-dependent thermodynamic properties and mixture behavior for pump systems.

Community
Intermediate