Process Mass Intensity: How to Calculate It and Benchmark a Process Fairly

Two plants can report the same product at very different process mass intensities, and neither number need be wrong. One counted its wash water and the other did not, or one stopped at the wet cake and the other at dried product. Process mass intensity (PMI) is the total mass of materials fed to a process divided by the mass of product it delivers. The figure means little until its boundary and basis are stated. Every kilogram counted was bought, charged, heated, separated and then recovered or treated, so PMI points straight at plant cost. This guide shows how to calculate process mass intensity, which decisions change the answer, what published benchmarks cover and how to compare a process fairly. It follows the method of the ACS Green Chemistry Institute Pharmaceutical Roundtable and peer-reviewed benchmarking data, within our chemical reaction engineering practice.

What Process Mass Intensity Measures

Process mass intensity answers a plain question: how many kilograms of material go into the plant for every kilogram of product that comes out? It counts everything that enters the process, so solvents, water, reagents and process aids weigh in alongside the reactants. That makes it a direct reading of how much material a plant buys, handles, heats, separates and eventually treats per unit of product. The ACS Green Chemistry Institute Pharmaceutical Roundtable defines it, and its definition is the one most benchmarks use.

The Roundtable's published definition has three parts, set out in its 2011 PMI tool launch presentation and its 2014 Convergent PMI Calculator presentation:

• Process — all steps of a synthetic path from commonly available materials to the final bulk active pharmaceutical ingredient (API).
• Raw materials — all materials, including water, used directly in synthesising, isolating and purifying the final form.
• Bulk API out — the final form of the active ingredient, dried to the expected specification. For other products, read it as the dried product at specification.

Written as a ratio, the process mass intensity formula is the total mass of inputs divided by the mass of dried product, in kg/kg. A value of 1 would mean every kilogram fed leaves as product. Multistep syntheses sit far above it, because solvents, water and excess reagents are not part of the molecule. When the E-factor counts every non-product stream as waste and uses the same boundary, waste equals inputs minus product, so PMI equals the E-factor plus one.

Jimenez-Gonzalez, Ponder, Broxterman and Manley made the case for PMI in pharmaceutical development in “Using the Right Green Yardstick”, published in 2011 in Organic Process Research & Development. The PMI metric has since spread from small-molecule APIs to biologics, peptides and oligonucleotides, as the benchmarks below show.

The definition is short, but a useful number needs a real mass balance behind it. The next section works one through with round numbers.

A Worked Example: Reading Process Mass Intensity by Stream

A process mass intensity calculation is most useful when the result is split by stream, because the split shows where the mass goes. The example below is hypothetical arithmetic for one batch step, not a real process and not a benchmark. Round numbers keep the method visible. Everything charged to the batch is listed, then divided by the dried product.

The total charged mass of 1,307 kg over 100 kg of dried product gives a PMI of 13.1 kg/kg. Water and the three solvent streams supply 12.0 of that, so trimming reagent charges would barely move the figure here. The same inputs give different answers when the boundary changes:

• Excluding water — (1,307 − 520) / 100 = 7.9 kg/kg. The Roundtable's calculator also reports a cumulative PMI excluding water.
• Net of solvent recovery — if 70% of the 540 kg of reaction and wash solvent is distilled and reused in the next batch, 378 kg is recovered. Net PMI is (1,307 − 378) / 100 = 9.3 kg/kg.
• Lower yield — the same inputs delivering only 80 kg of dried product give 1,307 / 80 = 16.3 kg/kg.
• Waste view — on the gross basis including water, the E-factor is 13.07 − 1 = 12.1 kg of waste per kg of product.

One set of inputs therefore yields three defensible figures, 13.1, 7.9 and 9.3 kg/kg, depending on how water and recovered solvent are treated. A yield loss then moves the first of them to 16.3 kg/kg. Each is correct for its stated boundary, which is the whole case for stating it. The next section lists the boundary decisions to settle first.

Setting the System Boundary Before Comparing PMI

Most disputes over a PMI value are disputes about the boundary, not the arithmetic. The same mass balance yields different answers depending on which steps, streams and recycles are counted. Writing the boundary down before collecting data is the cheapest way to make a comparison defensible. Six decisions matter most.

1. Start point — the Roundtable counts from commonly available materials. Starting from an advanced intermediate leaves out the mass spent making it, so a short route can look leaner than it is.
2. End point — dried product at specification, not a solution, wet cake or crude.
3. Water — the Roundtable definition includes it. PMI with water vs without water is a large swing. In the Roundtable's 2008 pharmaceutical benchmarking, water made up 28% of PMI, so excluding it removes more than a quarter of the figure.
4. Recovered solvent — gross PMI counts every kilogram charged, while net PMI counts only fresh input. Gross vs net solvent input was the largest swing in the worked example.
5. Cleaning and rinse materials — equipment rinses and cleaning solvents often sit outside a recipe-based figure. Benison and Payne's 2022 review, “Manufacturing mass intensity”, follows the metric through its first 15 years. It extends the metric into plant cleaning under the name manufacturing mass intensity.
6. Quantity basis — use charged masses from batch records, not recipe targets, and split solutions of intermediates into solute and solvent so nothing is counted twice.

A boundary written on one page and agreed before data collection settles most later arguments. It also fixes the denominator, because the dried product mass is what every wash, filtration and drying loss comes out of. Multistep routes add one more layer, which is how PMI accumulates from step to step.

Step, Cumulative and Convergent PMI

Real syntheses have several steps, so PMI exists at more than one level. Step PMI describes a single operation, and cumulative PMI follows the whole route back to commonly available materials. Convergent process mass intensity applies where separately made fragments are joined, and each branch is tracked on its own. The distinction decides where improvement effort goes.

The Roundtable released a free linear PMI calculator in 2011. It reports step PMI and cumulative PMI, with separate figures for solvents, water and reagents and a cumulative value excluding water. The Convergent PMI Calculator followed in 2014. It accepts up to 11 steps in a linear sequence and up to three branches of 11 steps each, with further branching possible.

Linear vs convergent PMI matters because a convergent route makes its fragments in parallel, and the final figure sums the mass behind every branch. Step PMI shows which operations contribute most to the cumulative value. A step with a high step PMI carries a large solvent and water load through the filters, dryers and recovery equipment, so it deserves attention first. For planning, the Roundtable offers the PMI Predictor, a web app that estimates PMI based on the phase of development. Borovika and co-authors describe it in Nature Sustainability in 2019.

With the method settled, the next question is what the published numbers can and cannot tell you.

Process Mass Intensity Benchmarks and What They Cover

A benchmark helps only when its boundary matches yours. The best-documented published PMI data come from the pharmaceutical industry, where the Roundtable has run benchmarking exercises since 2008. They show how wide the range is across product types. They do not give a target for a surfactant, resin or coating intermediate.

The first table uses Roundtable PMI figures as quoted by Kekessie and co-authors in their peptide manufacturing analysis in the Journal of Organic Chemistry in 2024. Each row is a different product class, and the paper quotes them side by side.

The same sources show where the mass sits. In the Roundtable's 2008 pharmaceutical benchmarking, organic solvents made up 58% of PMI, water 28%, reactants 8% and other process materials 6%. Kekessie and co-authors put the solvent contribution at 80 to 90% for small molecules. They report water at 28% of PMI for small-molecule processing, against 44% for solid-phase peptide synthesis and 51% for oligonucleotides, and link the higher shares to chromatographic purification.

The second table gives a wider view in kilograms of waste per kilogram of product. It shows Roger Sheldon's E-factor ranges by industry sector, as reproduced in Wikipedia's article on green chemistry metrics.

Benchmarks show the range, and your own baseline sets the target. PMI is also only one of several metrics, so the next section compares it with its neighbours.

PMI vs E-Factor vs Life Cycle Assessment

PMI, the E-factor and life cycle assessment answer different questions, so the choice between them follows the purpose. PMI and the E-factor count mass at the plant boundary. Life cycle assessment (LCA) follows environmental impacts through the supply chain and the product life, at a higher cost in data and time. The table sets them side by side.

PMI vs life cycle assessment is a question of what the decision needs. If the question is how much material the plant moves per kilogram of product, PMI answers it. If the question is which route has the smaller impact on carbon, water or toxicity, mass alone cannot rank the options and LCA is the tool.

PMI vs E-factor is mostly a matter of convention, since the two differ by one for the same boundary. Our guide to green chemistry metrics compares them with atom economy and reaction mass efficiency. Details of the PMI-LCA tool are on the Roundtable's PMI Life Cycle Assessment page, and the standards behind full LCA are summarised in Wikipedia's life cycle assessment article.

Sustainability Benchmarking: Six Steps to a Fair Comparison

Fair benchmarking is mostly bookkeeping. Two PMI values can be compared only when they share a boundary, a basis and a standard of data quality. The six steps below put that into sequence, and the first three carry most of the weight. They work for a single plant tracking its own progress and for a comparison against published figures.

1. Write the boundary — start point, end point, treatment of water, treatment of recovered solvent and cleaning scope, on one page.
2. Take charged masses from batch records — use what was weighed into the vessel, including repeat washes and rework, not the recipe target.
3. Divide by dried product at specification — use the isolated mass after filtration, drying and sampling losses.
4. Split by stream — reactants, reagents, solvents, water and process aids. The Roundtable asks suppliers to include this breakdown with the PMI data they provide.
5. Compare like with like — same product class, similar scale and the same development stage, using external numbers only where their boundary is known.
6. Track against your own history — one figure per product family and campaign. The Roundtable lists quantifying improvement through the development life cycle as a reason to measure PMI, and a falling value on a constant boundary shows real progress.

When a complexity-adjusted comparison is wanted, the Green Aspiration Level addresses it. Roschangar, Sheldon and Senanayake introduced it in 2015 to quantify the relative greenness of a production process while taking the complexity of its synthesis into account.

Laboratory PMI is a prediction of the plant figure, and it holds only as far as the lab procedure resembles plant operation. Our guides to scale-down models and pilot plant sizing cover how far bench data can be trusted and when a pilot run is needed. A pilot run also gives a mass balance at equipment scale, with real wash volumes and real losses.

Troubleshooting a High Process Mass Intensity

A high process mass intensity is a symptom with a short list of causes, and the stream split usually points at one of them. Start with the largest contributor and work down. The table pairs common patterns with the stream where the mass enters and the first check to make. Each fix has a plant consequence, either in the solvent recovery load or in the effluent bill.

Crystallisation sets antisolvent volume, wash ratio and mother liquor losses, all of which feed straight into PMI. Our guide to cooling, antisolvent or reactive crystallisation explains how route choice changes them.

A stream split turns a high PMI from a verdict into a work list. The remaining question is what the number cannot tell you.

What Process Mass Intensity Cannot Tell You

PMI is a mass ratio, and a mass ratio gives a kilogram of water and a kilogram of a hazardous solvent the same weight. That is its strength as a simple tracking metric and its weakness as a sustainability claim. The Roundtable's own authors state the limits plainly. Knowing them prevents a common misuse of the number.

• Material type — PMI does not take into account the type of material. Kekessie and co-authors describe DMF, the dominant organic solvent in solid-phase peptide synthesis, as highly hazardous and record its restriction under EU REACH. In PMI it weighs the same as water.
• Energy and logistics — neither is counted, so a change that cuts solvent but adds heating shows only the gain in the figure.
• Upstream impact — PMI does not account for the environmental impact incurred in making the starting building blocks and reagents.
• Starting-material complexity — a route that buys an elaborate building block can show a low PMI, because the mass spent making it sits outside the boundary.

Used within these limits, PMI is the most practical tracking metric a plant can run on its own batch records. Hazard review and LCA then cover what mass cannot. The last section turns this into a decision framework.

Which Metric to Track at Each Stage, and When to Get Help

Choosing the metric is a decision about project stage more than preference. Early on there is no mass balance, so design metrics and predictions do the work. Once batch data exist, PMI split by stream becomes the working number. When a decision depends on impact rather than mass, the question moves to LCA.

1. Route on paper — atom economy, reaction mass efficiency and a predicted PMI.
2. First lab batches — PMI by stream from charged masses, with the boundary written down.
3. Pilot or campaign data — PMI from equipment-scale batches. Rank the streams, then test solvent recovery and wash changes against the baseline.
4. Commercial running — PMI per campaign, plus cleaning materials as manufacturing mass intensity.
5. External claims and route choices with impact trade-offs — LCA to ISO 14040 and ISO 14044, with the method stated.

A process sustainability review turns this sequence into numbers for a specific plant. It delivers a written boundary, a stream-split mass balance from batch records and a ranked list of levers: solvent recovery design, wash optimisation and a waste minimisation study. Plants without a team for this can commission the first baseline from a process sustainability consultant or a green chemistry consultant, then run it internally. The same work extends to an emissions and effluent strategy. Every kilogram of input that does not leave as product ends up in a recovery unit, an emission or a waste stream.

Where the review leads to route or equipment changes, the team can carry them through as contract process engineering and process development services, with scale-up support for the plant transfer. Manufacturers in India and the Gulf who supply multinational buyers can use the same baseline to answer sustainability questionnaires with a stated boundary instead of a round number. The broader context is in our guide to eco-friendly chemical products, and the engagement starts with our process consulting service.