<?xml version="1.0" encoding="UTF-8" ?><!-- generator=Zoho Sites --><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><atom:link href="https://lab.wittigonia.net/log/tag/simulation/feed" rel="self" type="application/rss+xml"/><title>WITTIGONIA Lab - Log #Simulation</title><description>WITTIGONIA Lab - Log #Simulation</description><link>https://lab.wittigonia.net/log/tag/simulation</link><lastBuildDate>Thu, 19 Nov 2026 03:38:13 +0100</lastBuildDate><generator>http://zoho.com/sites/</generator><item><title><![CDATA[Equilibrium Check for System Dynamics Models]]></title><link>https://lab.wittigonia.net/log/post/equilibrium-check-model-quality-assurance</link><description><![CDATA[<img align="left" hspace="5" src="https://lab.wittigonia.net/Blog Images/saas-pipeline-stock-and-flow-structure.png"/>How to test a system dynamics model at rest: start in equilibrium, check consistency and conservation, then test credibility. SaaS pipeline example.]]></description><content:encoded><![CDATA[<div class="zpcontent-container blogpost-container "><div data-element-id="elm_M4xDy4TMQUmKgoeQkgAVWw" data-element-type="section" class="zpsection "><style type="text/css"></style><div class="zpcontainer-fluid zpcontainer"><div data-element-id="elm_a81GWpSVSECxO2VqHQNkug" data-element-type="row" class="zprow zprow-container zpalign-items-flex-start zpjustify-content- " data-equal-column="false"><style type="text/css"></style><div data-element-id="elm_ESn5dfXFQOqKF-O5RtbEiQ" data-element-type="column" class="zpelem-col zpcol-12 zpcol-md-12 zpcol-sm-12 zpalign-self- "><style type="text/css"></style><div data-element-id="elm_bdedoaYwLvHPmxVWUd61lg" data-element-type="codeSnippet" class="zpelement zpelem-codesnippet "><div class="zpsnippet-container"><article class="wl-note wl-note--a" lang="en" data-wl-template="lab-note" data-wl-version="0.4"><div class="wl-grid"><div class="wl-rail"><dl class="wl-meta"><div class="wl-meta__id"><dt>Lab note</dt><dd>LN-2026-005</dd></div>
<div><dt>Published</dt><dd><time datetime="2026-11-15">15 Nov 2026</time></dd></div>
<div><dt>Status</dt><dd><span class="wl-status wl-status--complete">● Complete</span></dd></div>
<div><dt>Reading time</dt><dd>6 min</dd></div><div><dt>Category</dt><dd><a href="https://lab.wittigonia.net/log/simulation">Simulation</a></dd></div>
</dl><nav class="wl-toc" aria-label="On this note"><p class="wl-toc__t">On this note</p><ol><li><a href="#a-three-stock-saas-pipeline-is-enough-to">A three-stock SaaS pipeline is enough to test the method</a></li><li><a href="#in-equilibrium-every-stock-has-inflow">In equilibrium, every stock has inflow equal to outflow</a></li><li><a href="#consistency-a-balanced-model-stays-flat">Consistency: a balanced model stays flat</a></li><li><a href="#conservation-what-flows-in-must-flow-out">Conservation: what flows in must flow out or stay in</a></li><li><a href="#credibility-the-response-must-be-one-we">Credibility: the response must be one we can defend</a></li><li><a href="#policy-test-onboarding-time-moves-the">Policy test: onboarding time moves the level, trial time only delays it</a></li><li><a href="#frequently-asked-questions">Frequently asked questions</a></li></ol></nav></div>
<div class="wl-main"><p class="wl-dek">We check a simulation model at rest before we trust it for analysis or policy testing. This note walks through the method with a small SaaS pipeline, for modellers who want a yes/no quality test.</p><div class="wl-tldr" role="note" aria-label="Summary"><p class="wl-tldr__t">TL;DR</p><p>We trust a model for analysis and policy testing only after it passes a quality check at rest. We start it in equilibrium, confirm it stays there, confirm that nothing is created or lost, and confirm it responds in a way we can defend when we change it. A model that drifts with no input has a structural or initialisation problem, and we want to find it before anyone sees a result.</p><ul><li>A units check proves the equations balance dimensionally. An equilibrium check proves the model behaves consistently.</li><li>Run both before any analysis or policy test.</li></ul></div>
<section class="wl-sec" id="a-three-stock-saas-pipeline-is-enough-to" aria-labelledby="a-three-stock-saas-pipeline-is-enough-to-h"><p class="wl-kicker">// The example</p><h2 id="a-three-stock-saas-pipeline-is-enough-to-h">A three-stock SaaS pipeline is enough to test the method</h2><p>New trials enter the pipeline, move through onboarding and end up as active customers in operations. The model has three stocks and these parameters:</p><ul><li>100 new trials start per week.</li><li>Trial lasts 2 weeks. 40% move on to onboarding, 60% leave.</li><li>Onboarding lasts 3 weeks. 90% move on to operations, 10% leave.</li><li>Customers in operations churn at 2% per week.</li></ul></section><section class="wl-sec" id="in-equilibrium-every-stock-has-inflow" aria-labelledby="in-equilibrium-every-stock-has-inflow-h"><p class="wl-kicker">// Equilibrium</p><h2 id="in-equilibrium-every-stock-has-inflow-h">In equilibrium, every stock has inflow equal to outflow</h2><figure class="wl-fig"><img src="https://lab.wittigonia.net/Blog%20Images/saas-pipeline-stock-and-flow-structure.png" alt="Stock and flow diagram of a SaaS customer pipeline with three stocks: Trial, Onboarding and Operations. Customers enter Trial through the converting flow, driven by the start rate. They move to Onboarding through buying, which depends on trial time and the fraction onboarding. They move on to Operations through becoming productive, which depends on onboarding time and the fraction retained. Each stock loses customers through its own churn rate." width="1600" height="900" loading="lazy" decoding="async"><figcaption><b>Fig 1</b>Stock and flow structure of the SaaS pipeline: customers move from trial to onboarding to operations, and each stage has its own churn outflow.</figcaption></figure><p>For a pipeline stage, the starting value is throughput multiplied by the average time an item spends in the stage. For a stock that only loses items at a fractional rate, it is inflow divided by that rate. We use this as a check, not as a rule for every model: it is how we assure quality before a model is trusted for analysis.</p></section><section class="wl-sec" id="consistency-a-balanced-model-stays-flat" aria-labelledby="consistency-a-balanced-model-stays-flat-h"><p class="wl-kicker">// Consistency</p><h2 id="consistency-a-balanced-model-stays-flat-h">Consistency: a balanced model stays flat</h2><p>In our pipeline, the equilibrium stocks are:</p><ul><li>Trial: 100 per week x 2 weeks = 200 customers.</li><li>Onboarding: 40 per week x 3 weeks = 120 customers.</li><li>Operations: 36 per week / 0.02 per week = 1,800 customers.</li></ul><p>All three lines must run flat. If a stock moves before we apply any shock, something is wrong. Typical causes are a missing or mis-wired flow, or initial values that do not match the equations. The check turns &quot;does this look right?&quot; into a yes/no test.</p><p>The distribution also tells a story. 1,800 of the 2,120 customers sit in operations, and only 200 in trial. If that picture looked implausible against the real business, we would revisit structure or parameters first.</p><div class="wl-table"><table><caption>Units check and equilibrium check test different things</caption><thead><tr><th scope="col">Check</th><th scope="col">What it tests</th><th scope="col">Typical failure it finds</th></tr></thead><tbody><tr><th scope="row">Units check</th><td data-label="What it tests">Dimensional consistency: does every equation balance in its units, such as customers per week integrating into customers?</td><td data-label="Typical failure it finds">An equation whose units do not balance</td></tr><tr><th scope="row">Equilibrium check</th><td data-label="What it tests">Behavioural consistency: does the model stay at rest when nothing is changed?</td><td data-label="Typical failure it finds">A stock that drifts because of a wrong initial value or a missing flow</td></tr></tbody></table></div>
<p>Vensim reports a passing units check as &quot;Units are AOK&quot; (source 1). Units errors do not prevent a simulation (source 2), and a model can pass the units check and still be wrong.</p></section><section class="wl-sec" id="conservation-what-flows-in-must-flow-out" aria-labelledby="conservation-what-flows-in-must-flow-out-h"><p class="wl-kicker">// Conservation</p><h2 id="conservation-what-flows-in-must-flow-out-h">Conservation: what flows in must flow out or stay in</h2><p>There is no magic loss or gain of people. In our pipeline, 100 trials per week come in. Out go 60 lost in trial, 4 lost in onboarding and 36 churned from operations, which is also 100. At any time, initial total + cumulative inflow - cumulative outflow = sum of all stocks.</p><p>We also test whether the model stays balanced when its structure changes. We add an outflow: involuntary churn of 0.5% per week on top of the existing 2%. The operations equilibrium becomes 36 / 0.025 = 1,440. If the initial value is derived from the equations, the model starts at 1,440 and stays flat. If the old value of 1,800 was hard-coded, the stock slides down to 1,440 on its own. That slide looks like business behaviour, but it is a modelling artefact, and exactly the kind of error that damages trust later.</p></section><section class="wl-sec" id="credibility-the-response-must-be-one-we" aria-labelledby="credibility-the-response-must-be-one-we-h"><p class="wl-kicker">// Credibility</p><h2 id="credibility-the-response-must-be-one-we-h">Credibility: the response must be one we can defend</h2><p>Only now do we trigger the system. We raise trial starts by 10%, from 100 to 110 per week, and change nothing else. Trial and onboarding settle within weeks. Operations needs well over a year to reach its new level of 1,980, because the 2% churn gives that stock a time constant of 50 weeks. That slow tail is the model&#x27;s own dynamics. It is not mixed up with a startup transient, so it can be read and explained.</p><p>We then move churn from the balanced state to 4%, then to 20%. The operations stock falls to 900, then to 180, following inflow divided by churn. We check that the response is proportional where the structure says it should be, that no stock goes negative, and that the numerics stay stable. We only see this non-linear response clearly because we started in balance.</p><p>A model that passes these checks is easier to believe and easier to use. When a stakeholder changes churn in a workshop, they see a response they can explain, not an unexplained line moving.</p></section><section class="wl-sec" id="policy-test-onboarding-time-moves-the" aria-labelledby="policy-test-onboarding-time-moves-the-h"><p class="wl-kicker">// Policy test</p><h2 id="policy-test-onboarding-time-moves-the-h">Policy test: onboarding time moves the level, trial time only delays it</h2><p>With the balanced model we can test operating policies. The question: how do different trial and onboarding durations change the number of customers in onboarding?</p><figure class="wl-fig"><img src="https://lab.wittigonia.net/Blog%20Images/saas-pipeline-policy-test-simulation.png" alt="Line chart of customers in onboarding over 52 weeks. All four runs start at 120. From week 10, the onboarding 6 weeks run and the trial 4 weeks plus onboarding 6 weeks run rise to a new equilibrium near 240. The combined run rises more slowly at first. The trial 4 weeks run dips to 94 at week 13 and returns to 120. The baseline stays at 120." width="1600" height="900" loading="lazy" decoding="async"><figcaption><b>Fig 2</b>Operating policy test: a longer onboarding duration raises the number of customers in onboarding from 120 to a new equilibrium near 240. A longer trial only causes a temporary dip.</figcaption></figure><p>Try it yourself: the model below uses the same parameters as the example in this note. Change a setting and watch how the pipeline responds.</p><wids-atom config="saas-journey" heading-level="3"><svg viewBox="0 0 640 350" width="100%" role="img" aria-label="Chart of a business in balance: 200 customers in trial, 120 in onboarding and 1,800 in operation, unchanged for two years." style="max-width:640px;height:auto;display:block;font:13px / 1.2 inherit;font-family:inherit;color:inherit;"><title>SaaS journey</title><text x="0" y="32" font-weight="600">In trial and onboarding</text><line x1="54" x2="440" y1="160.0" y2="160.0" stroke="currentColor" stroke-opacity=".14"/><text x="46" y="164.0" text-anchor="end" fill-opacity=".78">0</text><line x1="54" x2="440" y1="132.5" y2="132.5" stroke="currentColor" stroke-opacity=".14"/><text x="46" y="136.5" text-anchor="end" fill-opacity=".78">250</text><line x1="54" x2="440" y1="105.0" y2="105.0" stroke="currentColor" stroke-opacity=".14"/><text x="46" y="109.0" text-anchor="end" fill-opacity=".78">500</text><line x1="54" x2="440" y1="77.5" y2="77.5" stroke="currentColor" stroke-opacity=".14"/><text x="46" y="81.5" text-anchor="end" fill-opacity=".78">750</text><line x1="54" x2="440" y1="50.0" y2="50.0" stroke="currentColor" stroke-opacity=".14"/><text x="46" y="54.0" text-anchor="end" fill-opacity=".78">1,000</text><text x="0" y="204" font-weight="600">In operation</text><line x1="54" x2="440" y1="294.0" y2="294.0" stroke="currentColor" stroke-opacity=".14"/><text x="46" y="298.0" text-anchor="end" fill-opacity=".78">1,400</text><line x1="54" x2="440" y1="262.0" y2="262.0" stroke="currentColor" stroke-opacity=".14"/><text x="46" y="266.0" text-anchor="end" fill-opacity=".78">1,600</text><line x1="54" x2="440" y1="230.0" y2="230.0" stroke="currentColor" stroke-opacity=".14"/><text x="46" y="234.0" text-anchor="end" fill-opacity=".78">1,800</text><line x1="54" x2="440" y1="138.0" y2="138.0" stroke="currentColor" stroke-width="2.5"/><line x1="54" x2="440" y1="146.8" y2="146.8" stroke="currentColor" stroke-width="2.5" stroke-dasharray="8 5"/><line x1="54" x2="440" y1="230.0" y2="230.0" stroke="currentColor" stroke-width="3"/><text x="450" y="132.0" font-weight="700">Trial 200</text><text x="450" y="160.8" font-weight="700">Onboarding 120</text><text x="450" y="234.0" font-weight="700">In operation 1,800</text><text x="54" y="336" fill-opacity=".78">Weeks 0 to 104. Normal, if nothing changes.</text></svg><p>Chart of a business in balance: 200 customers in trial, 120 in onboarding and 1,800 in operation, unchanged for two years. <a href="https://sim.wittigonia.net/atoms/saas-journey/">Open the full model</a></p></wids-atom><p class="wi-atom-credit" style="font-size:0.875em;opacity:0.85;margin:0.5em 0 0;">Interactive model: <a href="https://sim.wittigonia.net/atoms/saas-journey/">SaaS journey</a> v1.0 by <a href="https://wittigonia.net/service/system-dynamics-simulation/">WITTIGONIA Dynamics</a> (Thomas Wittig), &copy; 2026 WITTIGONIA. Compiled with <a href="https://sdeverywhere.org/">SDEverywhere</a> by <a href="https://www.climateinteractive.org/">Climate Interactive</a> (MIT licence); authored in <a href="https://vensim.com/">Vensim</a> by Ventana Systems. <a href="https://sim.wittigonia.net/atoms/saas-journey/#credits">All credits, licences and how to cite</a>.</p><script src="https://sim.wittigonia.net/atoms/saas-journey/v1.0.2/saas-journey.v1.0.2.js" defer></script><p>Onboarding stock equals throughput times duration, so doubling onboarding from 3 to 6 weeks doubles the stock. A longer trial delays arrivals once, then the same 40 customers per week arrive again.</p><p>Only a model that passes both the units check and the equilibrium check goes on to analysis and policy testing.</p></section><section class="wl-sec" id="frequently-asked-questions" aria-labelledby="frequently-asked-questions-h"><p class="wl-kicker">// FAQ</p><h2 id="frequently-asked-questions-h">Frequently asked questions</h2><div class="wl-faq"><div class="wl-faq__q"><h3>What is an equilibrium check for a simulation model?</h3><p>It is a test at rest. Every stock starts at a value where inflow equals outflow, and we confirm the model stays flat until we change something. Drift without input points to a structural or initialisation problem.</p></div>
<div class="wl-faq__q"><h3>How do you calculate the equilibrium value of a stock?</h3><p>For a pipeline stage, multiply throughput by the average time an item spends in it: 40 customers per week x 3 weeks = 120. For a stock that loses a fixed fraction per week, divide inflow by that fraction: 36 per week / 0.02 per week = 1,800.</p></div>
<div class="wl-faq__q"><h3>What is the difference between a units check and an equilibrium check?</h3><p>A units check tests whether every equation balances dimensionally. Vensim reports a pass as &quot;Units are AOK&quot; (source 1), and units errors do not prevent a simulation (source 2). A model with correct units can still drift because of a wrong initial value or a missing flow, and the equilibrium check finds that drift.</p></div>
</div></section><section class="wl-sec wl-sources" id="sources" aria-labelledby="sources-h" data-wl-block="sources"><p class="wl-kicker">// Sources</p><h2 id="sources-h">Sources</h2><ol><li><a href="https://vensim.com/documentation/20570.html">Vensim Help: Checking for Model Syntax and Units Errors</a><small>accessed 2026-11-15</small></li><li><a href="https://vensim.com/documentation/ref_units_check.html">Vensim Help: Units Check</a><small>accessed 2026-11-15</small></li></ol></section><aside class="wl-cta" aria-labelledby="cta-t"><p class="wl-cta__t" id="cta-t">Testing a model of your own?</p><p>We run management simulation workshops where teams build and test models like this one.</p><div class="wl-cta__acts"><a class="wl-btn" href="https://wittigonia.net/service/management-simulation-workshops/">Explore management simulation workshops</a></div>
</aside><aside class="wl-author" aria-labelledby="author-t"><p class="wl-author__t" id="author-t">About the author</p><p><a href="https://wittigonia.net/about/team/thomas-wittig/">Thomas Wittig</a> is CEO and founder of WITTIGONIA. In the <a href="https://wittigonia.net/service/system-dynamics-simulation/">WITTIGONIA dynamics</a> practice he builds simulation models for policy testing and facilitates management simulation workshops with enterprise teams.</p></aside><footer class="wl-end"><ul class="wl-tags" aria-label="Tags"><li><a href="https://lab.wittigonia.net/log/tag/simulation">Simulation</a></li><li><a href="https://lab.wittigonia.net/log/tag/system-dynamics">System Dynamics</a></li><li><a href="https://lab.wittigonia.net/log/tag/saas">SaaS</a></li></ul><p class="wl-wai"><picture class="wl-wai__mark"><source srcset="/WI%20Brand/wai-mark-dark.webp" media="(prefers-color-scheme: dark)"><img src="/WI%20Brand/wai-mark-light.webp" alt="WAI, WITTIGONIA AI declaration" width="62" height="18" decoding="async"></picture><span>AI-assisted · Human-reviewed</span><a href="https://wittigonia.net/ai/">Learn more at wittigonia.net/ai</a></p></footer></div>
</div><script type="application/ld+json">{"@context":"https://schema.org","@graph":[{"@type":"BlogPosting","@id":"https://lab.wittigonia.net/log/post/equilibrium-check-model-quality-assurance#note","headline":"Equilibrium Check for System Dynamics Models","identifier":"LN-2026-005","datePublished":"2026-11-15","dateModified":"2026-11-15","articleSection":"Simulation","keywords":["equilibrium check","system dynamics model validation","model quality assurance","units check"],"image":"https://lab.wittigonia.net/Blog%20Images/saas-pipeline-stock-and-flow-structure.png","author":{"@type":"Person","name":"Thomas Wittig","url":"https://wittigonia.net/about/team/thomas-wittig/"},"publisher":{"@type":"Organization","name":"WITTIGONIA GmbH","url":"https://wittigonia.net/"},"mainEntityOfPage":"https://lab.wittigonia.net/log/post/equilibrium-check-model-quality-assurance","timeRequired":"PT6M","hasPart":{"@type":"WebApplication","@id":"https://sim.wittigonia.net/atoms/saas-journey/#app","name":"SaaS journey","description":"Interactive simulation of a three-stock SaaS customer pipeline (trial, onboarding, operation) that starts in equilibrium. Change the trial length and the onboarding length from week 10 and watch how each stage responds over 104 weeks.","url":"https://sim.wittigonia.net/atoms/saas-journey/","applicationCategory":"EducationalApplication","operatingSystem":"Web browser","browserRequirements":"Requires HTML5 and JavaScript","featureList":["Slider for the trial length (2 to 10 weeks)","Slider for the onboarding length (3 to 15 weeks)","Four preset scenarios: Baseline, Longer trial, Slower onboarding, Both","Charts of customers in trial, onboarding and operation over 104 weeks, with the change starting at week 10","Static chart and text fallback without JavaScript"],"keywords":["system dynamics","equilibrium check","SaaS customer pipeline"],"softwareVersion":"1.0.2","inLanguage":"en","isAccessibleForFree":true,"offers":{"@type":"Offer","price":"0","priceCurrency":"EUR"},"creator":{"@type":"Person","name":"Thomas Wittig","url":"https://wittigonia.net/about/team/thomas-wittig/"},"provider":{"@type":"Organization","name":"WITTIGONIA GmbH","url":"https://wittigonia.net/"}}},{"@type":"FAQPage","mainEntity":[{"@type":"Question","name":"What is an equilibrium check for a simulation model?","acceptedAnswer":{"@type":"Answer","text":"It is a test at rest. Every stock starts at a value where inflow equals outflow, and we confirm the model stays flat until we change something. Drift without input points to a structural or initialisation problem."}},{"@type":"Question","name":"How do you calculate the equilibrium value of a stock?","acceptedAnswer":{"@type":"Answer","text":"For a pipeline stage, multiply throughput by the average time an item spends in it: 40 customers per week x 3 weeks = 120. For a stock that loses a fixed fraction per week, divide inflow by that fraction: 36 per week / 0.02 per week = 1,800."}},{"@type":"Question","name":"What is the difference between a units check and an equilibrium check?","acceptedAnswer":{"@type":"Answer","text":"A units check tests whether every equation balances dimensionally. Vensim reports a pass as \"Units are AOK\", and units errors do not prevent a simulation. A model with correct units can still drift because of a wrong initial value or a missing flow, and the equilibrium check finds that drift."}}]}]}</script></article></div>
</div></div></div></div></div></div> ]]></content:encoded><pubDate>Sun, 15 Nov 2026 09:52:00 +0100</pubDate></item></channel></rss>