Potency/Assay Adjustment – Recipe Compensation

This topic is part of the SG Systems Global regulatory & operations glossary.

Updated October 2025 • Assay/Potency, Moisture & Titer, Base Equivalents • R&D, Manufacturing, QA/RA, Lab, Supply Chain

Potency/assay adjustment is the governed method of changing the as‑received addition mass of an ingredient so the batch still receives the intended active amount. It converts real‑world assay, moisture, or activity into a precise addition that preserves the recipe’s science—no wishful rounding, no “close enough.” If the label claim, efficacy, critical quality attributes (CQAs), or safety hinge on actives (API, enzyme, vitamin, colorant, preservative, fragrance allergen thresholds), then potency math is not optional; it’s the difference between right‑first‑time and rework. The blunt truth: if potency factors live in private spreadsheets or operator head‑math, you’re gambling with compliance and margin.

“Dose actives by truth, not by label: the drum’s assay pays the bill your recipe wrote.”

1) What Potency Adjustment Covers—and What It Does Not

Covers: corrections for assay/purity, water/solvent content (loss‑on‑drying), base‑equivalent conversions (salt ↔ free acid/base), biological/enzymatic activity units to mass, titer for solutions, and potency decay over time. It includes UOM conversions, rounding policy, and how the MES computes and displays the exact addition with windows for alert/action limits. It also covers supplier variability, incoming testing, component release, and linkages to labels and claims.

Does not cover: changing efficacy claims by stealth or ignoring regulatory specifications. If potency is off‑spec, you don’t “compensate it into tolerance”; you trigger Deviation/CAPA and address the lot or the supplier. Potency math is not a permit to override the QMS.

2) Legal, System, and Data Integrity Anchors

Ingredients used in regulated manufacturing require documented identity, strength, quality, and purity. Pharma: 21 CFR 211/210. Food: 21 CFR 117 and your FSP/HACCP. Cosmetics: implement under ISO 22716 and ensure label accuracy (INCI). Electronic potency evidence must meet Part 11/Annex 11: validated apps (CSV/GAMP 5), unique users, audit trails, and governed retention. Bottom line: potency math belongs in controlled systems, not in ad‑hoc calculators.

3) The Evidence Pack for Potency Compensation

A credible “potency pack” includes: (i) recipe actives defined as mass/unit of active or base‑equivalent with windows; (ii) UOM and rounding rules under a governed service (UOM Conversion); (iii) supplier COA and in‑house assay results; (iv) lab method validation and capability (TMV, MSA, 17025); (v) equivalence math (salt‑to‑base, activity‑to‑mass) expressed as governed formulas; (vi) executed eBMR showing the exact computation and addition; (vii) SPC/CPV trends for assay drift; (viii) deviations/CAPAs and supplier SCAR where needed. Auditors should rebuild your numbers in minutes from first principles.

4) Core Math—From Active Requirement to Addition Mass

Define Required Active for the batch (e.g., 1.500 kg preservative solids). Obtain as‑received potency (e.g., 60% w/w) and moisture (e.g., 5% LOD if potency reported on dry basis). Compute a correction factor and the addition mass:

• Simple assay (as‑received): Addition = RequiredActive ÷ Assay_{as‑received}.
• Assay on dry basis (db) with LOD: Assay_{as‑received} = Assay_db × (1 − LOD). Then compute as above.
• Salt ↔ base equivalents: RequiredActive (as base) × (MW_salt ÷ MW_base) to get salt‑equivalent, then divide by assay.
• Activity units: RequiredActive (IU or U) ÷ (IU or U per gram) = grams to add.

Do calculations in high precision, round once at presentation. Bake these formulas into the MES—not into somebody’s desktop workbook.

5) As‑Received, Anhydrous, and “db” Confusion—Kill Ambiguity

Vendors report assay variously: as‑is, anhydrous, solvent‑free, or “dry basis.” Your system must store the basis with the value and convert deterministically. Keep a canonical attribute for Assay Basis on each lot, tied to the math used at execution. If Assay Basis changes via supplier NoC, route via Change Control and re‑validate the arithmetic. Assay is a number with context; without context it’s a rumor.

6) Incoming Testing—Trust, But Verify

Goods Receipt and Incoming Inspection should confirm COA potency before Component Release. Decide risk‑based sampling frequency and acceptance criteria under your QMS; prove the method works (TMV) and the lab is competent (17025). If vendor drift or lot spread is material, upgrade the supplier or raise a SCAR. Your batch should not be the first time you discover potency fiction.

7) Lab Methods—Get the Signal, Not the Noise

Method validation must cover specificity, accuracy, precision, range, and robustness for the matrix at hand (TMV). Run MSA to quantify analytical repeatability/reproducibility and separate measurement noise from material variability. If R&R eats half your window, your process is not “variable”—your measurement is. Fix the method or widen the window honestly under Change Control.

8) Micro vs. Macro—Precision Where It Hurts

Potency‑corrected targets often land in the gram to tens‑of‑grams range for powerful actives. Weigh them with the discipline in Micro‑Ingredient Dosing (weigh‑by‑difference, minimum weight, anti‑static) and drive bulk carriers in Macro Dosing. Let the MES split the adjusted target across steps—don’t ask operators to “trim” off the cuff. If you can’t hit the adjusted micro within window, your balance or method is under‑capable—fix physics, not people.

9) Solutions, Titer & Density

Liquid actives specified by titer (e.g., 40% w/w) or by volume percent demand a density‑aware approach. Store titer as w/w and the density table at reference temperature under Document Control. Dose by mass where possible; if dosing by volume, apply temperature‑corrected density (see UOM and density discipline). Tie setpoints to SCADA inputs so the MES does the math, not the person at the pump.

10) Units of Activity—IU, U, and Everything In Between

Enzymes, vitamins, and biologicals use activity units. Convert RequiredActive (IU or U) to grams via the lot’s certified IU/g (or U/g). If activity is temperature‑sensitive or time‑decaying, record lot age and storage conditions; apply decay only if justified and approved. Don’t invent “activity compensation” on the floor. If the activity is off‑spec, stop and escalate; do not quietly double the mass to “hit the feel.”

11) Salts, Hydrates & Free Base/Acid Equivalents

Some specifications are written on a free base/acid basis while materials ship as salts or hydrates. Store molecular weights and water of crystallization with the item master; compute base‑equivalent addition deterministically. Embed the equivalence formula in the platform alongside potency. If labels express actives as base, ensure Labeling Control pulls the same computation. Two versions of the truth create recalls.

12) Rounding & Significant Figures—Decide Once

Perform potency math at high precision; round only when presenting targets on screens/labels. Define significant figures per class (macro vs. micro) in the UOM conversion service. Ban hand rounding, double rounding, and inline calculator edits. When an exception is genuinely needed, force reason codes and e‑signatures recorded in the audit trail.

13) Workflow—From Receipt to Charge

1) Receive & verify: Vendor delivers COA; QA samples per sampling plans; lab confirms assay; WMS holds lot in quarantine. 2) Release: On pass, QA sets Component Release; assay basis/values are posted to master data. 3) Execute: MES computes addition mass from recipe active target + lot assay; Weigh/Dispense enforces windows; devices and users are attributable. 4) Label: If copy depends on actives, panels render from the same data; Label Verification scans confirm the variant. 5) Review: QA verifies eBMR calculations and assay usage; Release Status applied.

14) Labels, Claims & INCI—Say What You Do, Do What You Say

Ingredient panels (cosmetics), nutrition/fortification statements (food), and strength statements (drug) must reflect the same potency truth that drove execution. Drive artwork from governed masters (Labeling Control) and the INCI engine where applicable (INCI). Block print if potency or basis drift makes the panel inaccurate. “We’ll fix the label later” is how recalls are born.

15) Stability, Potency Decay & Hold Time

Some actives lose potency with time, temperature, light, or pH. Control storage conditions under Temperature Mapping and EM and set retest dates. Don’t assume linear decay; use stability studies for kinetic models. If you need to compensate for predicted decay (e.g., overages), document the basis and regulatory allowance in the recipe and label logic under Document Control. “Silent overage” is not a control strategy; it’s a finding waiting to happen.

16) SPC/CPV—Turn Drift Into Decisions

Trend assay by supplier, lot, season, and storage. Use SPC on assay values and Cpk on adjusted dosing error (executed minus target active). Feed into CPV. When drift is chronic, remove the compensation burden from operators by fixing supplier controls or by changing formulation to a more stable form—then lock changes via Recipe Versioning.

17) Supplier Governance—Fix the Source, Not the Symptom

Qualify suppliers (VQ/SQM), define potency specs and assay bases in purchase contracts, and monitor COA‑to‑lab alignment. Use SCAR for recurring drift or data integrity concerns. Procurement cannot “switch to a similar grade” without re‑approval; potency basis and activity units are not commodity trivia.

18) Data Integrity—Make Arithmetic Reconstructable

Every potency calculation must be attributable (who), legible (what), contemporaneous (when), original (where from), and accurate (how)—ALCOA(+). Put the math in validated systems (CSV), bind user/device identity (UAM), and enforce immutable audit trails. Test backups/restores (Record Retention). If a junior auditor can’t reproduce the addition mass from stored assay and formula in minutes, your controls are weak.

19) Mass Balance, Yield & Cost—Potency Shows Up in Finance

Potency compensation affects mass balance, yield variance, and cost per batch. Track the delta between nominal and potency‑adjusted addition by SKU and period; it’s real money. If you are consistently dosing 8% more preservative because vendors ship low assay, either negotiate price/assay or switch suppliers. “We knew it was a bit weak” is not a margin strategy; it’s a leak.

20) Common Pitfalls & How to Avoid Them

• Spreadsheet arithmetic. Move potency math into validated MES/eBMR with audit trails.
• Assay basis amnesia. As‑received vs. dry‑basis confusion. Store basis with the value and convert deterministically.
• Salt/base mistakes. Wrong molecular weights or hydrates. Govern equivalence data in master records.
• Ignoring moisture. LOD can dwarf assay precision. Include it or expect silent underdose.
• Rounded away actives. Early, repeated rounding. Round once at presentation.
• Label/execution mismatch. Labels rendered from free‑text while execution uses calculated bases. Drive both from one source.
• “Feel‑based” trims. No free‑pour. If trims exist, bound them with windows and e‑sign reasons.
• Unverified COAs. Trust but verify. Escalate with SCAR when patterns emerge.

21) Metrics That Prove Control

• COA vs. lab delta (%) by supplier and material.
• Adjusted‑dose variance (executed active − target active) with Cpk.
• Assay drift over time and time‑to‑SCAR closure.
• Rounding exception rate and reason‑code completeness.
• Label/eBMR parity incidents and time‑to‑correction.
• Overage cost vs. plan, tied to supplier lots.
• Audit‑trail review health (late entries, edits) for potency steps.

Good metrics cause decisions: supplier changes, method improvements, or recipe re‑baselining. Vanity metrics do not survive management review.

22) What Belongs in the Potency/Assay Record

Recipe active definition (mass, base‑equivalent, IU) and windows; UOM/rounding policy; molecular weights/hydrates and equivalence formulas; vendor COAs; in‑house lab results with TMV and MSA; assay basis and LOD; executed MES calculations with user/device IDs; label renderings; deviations/OOS/OOT with CAPA; supplier governance trail (VQ/SQM/SCAR); SPC/CPV charts; and approvals under Document Control.

23) How This Fits with V5 by SG Systems Global

Governed potency engine. The V5 platform stores assay values with basis (as‑received, dry basis), LOD, molecular weights/hydrates, IU/g, and density. It computes adjusted additions centrally and pushes exact targets to Weigh/Dispense, blocking steps on missing or off‑status data.

Device‑tight execution & evidence. V5 binds balances and feeders to steps, enforces windows, logs user/device identity, and writes the entire arithmetic chain into the eBMR with full audit trail. Labeling Control consumes the same truth to render compliant panels.

Supplier & quality loop. With WMS and EPCIS, V5 ties potency to genealogy; SPC/CPV dashboards surface assay drift and overage cost; CAPA and SCAR close the loop. Bottom line: V5 turns potency math from tribal knowledge into governed automation—defensible, repeatable, and cheaper.

24) FAQ

Q1. When must we adjust for potency?
Whenever the active requirement is defined independent of carrier (dry solids, base‑equivalent, IU) and the as‑received potency differs from nominal. If the CQA depends on the active, adjust—period.

Q2. How do we handle assay reported on a dry basis?
Convert to as‑received using LOD: Assay_{as‑received} = Assay_db × (1 − LOD). Then compute addition mass. Store basis and LOD with the lot and lock the formula in the MES.

Q3. What about salts/hydrates vs. free base/acid labels?
Convert using molecular weights (and waters of hydration) to base/acid equivalents, then divide by assay. Render labels from the same computation to avoid mismatch.

Q4. How often should we verify vendor assay?
Risk‑based. High‑impact actives and new suppliers merit every‑lot testing; stable, proven suppliers can be on reduced frequency with periodic verification. Trend deltas and escalate when drift emerges.

Q5. Can we use overages to cover stability loss?
Only if scientifically justified, documented in the recipe/label logic, and permitted by regulation or standard. Silent overages are not acceptable. Use stability data and route via Change Control.

Q6. What if potency is out‑of‑spec?
Don’t compensate it into tolerance. Quarantine the lot, investigate under Deviation, and disposition per QMS. Use SCAR if supplier‑related.

Related Reading
• Systems & Integrity: MES | eBMR | Audit Trail | UOM Conversion | Document Control
• Lab & Methods: TMV | MSA | ISO/IEC 17025 | Laboratory Tests
• Governance & Release: 21 CFR 211 | 21 CFR 117 | ISO 22716 | Release Status
• Operations & Scale: Micro‑Ingredient Dosing | Macro Dosing | Mass Balance | CPV | Cp/Cpk