The Discharge Monitoring Report is a legal document. Every number you submit is compared against your permit limits inside EPA's Integrated Compliance Information System (ICIS), and if a value exceeds a limit — or if a required value is missing — the system automatically generates a violation record without any human review. No grace period. No phone call first. The violation exists the moment ICIS runs its matching algorithm.
What surprises most operators is how many of those violations stem not from actual effluent problems, but from calculation errors that are entirely avoidable. The facility may have been in full compliance with its permit. The numbers just didn't reflect that because of how they were computed. Here are the seven mistakes we see most often, why they happen, and what to do instead.
Mistake 1: Using Arithmetic Mean for Bacteria Parameters
If your permit contains a monthly average limit for E. coli, fecal coliform, or total coliform, that limit almost certainly applies to the geometric mean — not the arithmetic mean. This is standard across NPDES permits for biological indicator organisms, and it is explicitly required by 40 CFR Part 136. Despite this, arithmetic averaging is one of the most common calculation errors we audit at small facilities.
The reason it happens is simple: operators use the same averaging method for everything. BOD, TSS, ammonia — arithmetic mean works fine for all of them. When bacteria results come in from the lab as colony-forming units per 100 mL, it feels natural to add them up and divide. The resulting number looks plausible. Nobody catches it until the next permit compliance review.
The geometric mean is calculated by multiplying all sample values together and taking the nth root, where n is the number of samples. Equivalently, you can take the average of the natural logarithms of each value, then raise e to that power. Most spreadsheet programs have a GEOMEAN() function that does this in one step.
The practical difference matters most when you have one very high outlier reading. Arithmetic averaging lets a single spike dominate the result and can push a facility over its limit even when most samples were fine. Geometric averaging dampens the effect of outliers — which is why regulators require it for pathogens, where the distribution of organisms in wastewater is log-normal by nature.
When ICIS compares your reported monthly average against your permit limit, it does not know or care how you computed the average. It just compares numbers. If you reported an arithmetic mean that exceeds the geometric mean limit, you have a D90 effluent violation in the system.
Mistake 2: Load Calculations Using Monthly Average Flow
Mass loading calculations — typically expressed as pounds per day — require you to multiply a concentration by a flow rate. The error occurs when operators use their monthly average daily flow for all load calculations rather than the flow measured on the same day as the sample.
Your permit's daily maximum load limit is based on worst-case daily conditions. Using the average monthly flow understates the actual load on high-flow days and can mask genuine exceedances. More importantly, it is the wrong methodology. The correct approach is: same-day measured flow times the concentration measured from that same sample event.
The conversion factor for a pounds-per-day calculation is: concentration (mg/L) × flow (MGD) × 8.34 (lb/gal). Each sample result needs to be paired with the flow recorded during that sample collection. When you then report a monthly average load, you average those per-sample load values — not the concentrations separately and flows separately.
This error is particularly common at facilities that receive flow data from a SCADA system and lab results from a separate email PDF, and manually reconcile them in a spreadsheet. The daily flow value used for one row gets accidentally carried into another. Auditing for this requires cross-referencing your sample log against your flow log by date.
Mistake 3: Wrong Non-Detect Substitution
When a lab result comes back as "less than the method detection limit" (MDL) — reported as something like <0.010 mg/L — you cannot enter zero on your DMR. You also cannot leave the cell blank. You need to substitute a value, and the substitution method matters.
The most common wrong approach is substituting zero. This artificially deflates averages and can actually cause problems in the opposite direction from what you'd expect: it makes your removal calculations look better than they are, which can mask analytical problems. More importantly, it is not the accepted method under EPA guidance.
The standard substitution under EPA's recommended approach is one-half the method detection limit (½ MDL). Some permits or state programs specify using the practical quantitation limit (PQL or MQL) instead, or ½ the PQL. Always check your permit and your state agency's guidance first. If your permit is silent on this, ½ MDL is the defensible default under EPA's current statistical guidance documents.
In practice: if your lab reports a non-detect for copper at <0.005 mg/L, enter 0.0025 on your DMR. Document this substitution in your calculation worksheet. When ICIS flags an apparent violation that was caused by incorrect non-detect handling, reconstructing what actually happened requires that documentation trail.
Mistake 4: Averaging pH Values Arithmetically
pH is a logarithmic scale. A pH of 6 is ten times more acidic than a pH of 7. A pH of 5 is one hundred times more acidic than a pH of 7. When you arithmetically average pH values, you are averaging the log of hydrogen ion concentrations — which produces a result that is mathematically meaningless and potentially wrong by a factor that could determine whether you are in or out of compliance.
Consider two samples: pH 5.0 and pH 9.0. Arithmetic average: 7.0. But the actual average hydrogen ion concentration corresponds to a pH of approximately 5.3, which is meaningfully different and could fall outside a permit range of 6.0–9.0.
The correct method is to convert each pH reading to hydrogen ion concentration using [H+] = 10^(-pH), average those concentrations, then convert back to pH using pH = -log10([H+] average). Many operators are not aware this is required because pH seems like a simple number. Labs typically report individual readings, not averages, so the conversion step falls to whoever is filling out the DMR.
In practice at most facilities, pH does not swing wildly enough for this distinction to trigger a violation. But in facilities treating industrial waste streams or with significant pH variability, the difference can matter — especially near the boundary of a 6.0 lower limit or a 9.0 upper limit. The calculation method is not optional: if your state inspector audits your calculation worksheet and sees arithmetic pH averaging, it is a methodology finding even if the numbers happened to come out within limits.
Worried about your calculations?
Effluent Labs audits your current DMR workflow and handles all calculations to EPA-standard methods. Free 15-minute consultation.
Talk to Us See How It WorksMistake 5: Rounding Errors on Percent Removal
Many secondary treatment permits include a percent removal requirement for BOD and TSS — typically 85% minimum. This limit exists alongside numeric effluent concentration limits, and both apply independently. You can meet your effluent concentration limit and still violate your percent removal limit, or vice versa.
The trap here is rounding. If your actual calculated percent removal is 84.94%, that rounds to 84.9% — which is below 85.0% and triggers an automatic violation in ICIS. If you round it to 85% before entering it, you are reporting a value you did not actually achieve, which is a separate problem.
The percent removal formula is: ((influent - effluent) / influent) × 100. The influent value used should be a representative composite sample from the same period as the effluent composite — not an estimate or a historical average. Small errors in your influent sampling can have an outsized effect on the percent removal calculation, especially when influent concentrations are low.
Seasonal variability is a common driver here. During wet weather dilution events, influent BOD concentrations can drop significantly. If both influent and effluent are low but the plant is removing a consistent absolute mass, the percent removal figure may dip below 85% mathematically even though the plant is performing normally. Document the conditions. If you have a bypass event or an unusual influent, note it on the DMR.
The fix is to carry your calculation to at least three decimal places throughout, report the actual computed value, and round only at the final entry step to the precision your permit specifies. Never round up to meet a limit.
Mistake 6: Selective Sample Reporting
Your permit specifies a monitoring frequency: once per week, twice per month, once per month. Every sample collected at the frequency and location specified in your permit must be reported on your DMR. You cannot omit a result because it was high. You cannot report only the best results from a month where you ran extra tests. Selective reporting — cherry-picking favorable results — is itself a violation of the permit, separate from any effluent exceedance.
This situation most often arises from good intentions. An operator sees an unexpectedly high result and runs a confirmatory sample. The confirmatory result is better. They report the confirmatory result and set aside the original. But the original sample was collected under the monitoring requirements of the permit, which means it is a required data point.
The correct procedure is to report all samples collected in accordance with your permit monitoring schedule, calculate your averages across all of them, and report any exceedances accurately. If you believe a sample result was caused by a sampling error or lab error, document that in writing and flag it with your regulatory agency — but still report the result. Many permits have provisions for requesting that an anomalous result be invalidated, but that process requires advance coordination, not after-the-fact omission.
Extra voluntary monitoring is fine and can be valuable. Results from voluntary extra samples can be noted separately in a comment field, but they do not replace required samples and cannot be used to average down required sample results unless your permit explicitly allows it.
Mistake 7: Using Wrong Permit Limits When a Permit Is Reissued Mid-Period
NPDES permits have five-year terms and are periodically reissued. When a new permit becomes effective mid-year, the limits that apply to each reporting period change as of the effective date of the new permit — not at the start of the next calendar year. Operators sometimes continue using limits from the expired permit for months after the new permit takes effect because they have not updated their DMR templates or calculation worksheets.
This is particularly consequential when the reissued permit includes new parameters, tighter limits, or changed monitoring frequencies. If your new permit adds a quarterly monitoring requirement for a parameter that was not previously required, you are responsible for having those samples collected starting from the effective date — even if no one called you to remind you.
The practical fix is to create a new DMR template the moment you receive your reissued permit and mark the effective date prominently. Review all changes against the prior permit: new parameters, changed limits (note whether limits changed for daily maximum, monthly average, or both), changed monitoring frequencies, and changed sample collection methods (composite vs. grab). If you use a third-party lab, notify them immediately of monitoring frequency changes so sample collection schedules are updated.
ICIS is loaded with your permit limits by your state NPDES authority as part of the permit issuance process. By the time your new permit takes effect, ICIS already has your new limits. If you report against old limits and ICIS is checking against new ones, every comparison is wrong. The system will generate violations — or worse, fail to generate violations for actual exceedances — based on the correct limits, not the ones you thought you were using.
The common thread across all seven mistakes is that they are process failures, not measurement failures. The underlying data was often correct. The error happened somewhere between the lab result and the submitted DMR value. Building a documented calculation workflow — with a second-person review before submission — catches most of these before they become ICIS violations.
If you are auditing your own DMR history, the best place to start is your ECHO record at echo.epa.gov. Search your facility by permit number and review the DMR submission history for any violation flags. For each flagged period, pull the original DMR and the permit limits in effect at that time and trace the discrepancy back to its source. Corrected DMRs can often be submitted in NetDMR to address historical calculation errors, though the window for doing so varies by state.