TCSC Water Quality Intelligence Series Pre-Treatment Intelligence · June 2026
The moment a sample is taken after treatment, you have a compliance record. The moment it is taken before treatment, you have a decision-making tool. The difference between those two positions is the difference between reporting and responding.
By Lesley Mukwada, M.Sc. · Founder & Director, The Chemistry Solutions Company (Pty) Ltd Partnered with SANS 241:2015 Accredited Water Quality Laboratory
Most water quality testing is designed to answer one question: did the treatment work? It is a necessary question. It is not the most useful one. The most useful question is asked upstream — before the chemistry changes, before the biology is disrupted, before the process has had a chance to transform what arrived. Pre-treatment data is not preparation for compliance. It is the earliest and most powerful signal in the entire water management chain.
This is a principle understood clearly in environmental engineering, in industrial hygiene, and in public health epidemiology. In water management, it is applied inconsistently. Utilities and treatment operators often invest heavily in effluent monitoring while treating influent characterisation as a preliminary formality. The result is a system that knows a great deal about its outputs and relatively little about its inputs — which is, analytically speaking, working backwards.
The Grab Sample Problem
The dominant methodology in water quality monitoring remains periodic grab sampling — a physical sample taken at a point in time, sent to a laboratory, and returned as a result that describes what the water was doing when it was collected. This is not a trivial limitation. It is a structural one.
Water quality is not static. Influent composition at a municipal treatment works varies with rainfall events, seasonal temperature shifts, upstream agricultural cycles, and the discharge patterns of industrial users in the catchment. A grab sample taken at 10:00 on a Tuesday may bear no relationship to what passed through the intake at 02:00 on Saturday. The grab sample is accurate. It is simply not representative.
“By the time a sample is processed, the process condition that created it has already passed.”— KETOS / Cincinnati MSD Case Study, 2024
This observation, made in the context of a large American municipal utility’s shift to continuous monitoring, captures a challenge that is as acute for South African water systems as it is for any other. The City of Cape Town’s distribution network, the municipal systems of the Western Cape’s agricultural interior, and the small water service authorities of the Northern Cape face the same epistemological problem: periodic data tells you what happened. Continuous pre-treatment data tells you what is coming.
What Influent Data Actually Reveals
A well-characterised influent sample — collected continuously, across a full parameter panel — carries information that post-treatment sampling cannot reconstruct. Consider what is present in raw water that treatment is specifically designed to remove or transform:
Ammonia (NH₃-N)Dominant nitrogen form in raw domestic and industrial wastewater. Converted to nitrate/nitrite by biological nitrification — invisible in effluent without influent baseline.
Natural Organic Matter (NOM)Humic and fulvic acids react with chlorine disinfectants to produce trihalomethanes (THMs) and haloacetic acids (HAAs) — the primary regulated disinfection by-products.
Turbidity & TSSSuspended particulate load drives coagulant dosing requirements and can shield pathogens from UV or chlorine disinfection.
pH & Alkalinity Pre-treatment pH determines optimal coagulation chemistry and chlorine speciation (HOCl vs OCl⁻). Alkalinity buffers the system through acid-producing biological processes.
COD / BOD Organic load drives oxygen demand in biological treatment. Industrial influent spikes in COD signal permit exceedance upstream — before the treatment plant absorbs the burden.
Heavy Metals: Industrial discharge of chromium, lead, and cadmium into collection systems can shock biological treatment processes and violate effluent limits for reasons entirely external to plant performance.
Each of these parameters undergoes transformation during treatment. Some are removed. Some are converted. Some combine with treatment chemicals to form new compounds. The only moment at which the raw, unmediated character of the source water is visible is before any of this occurs. That moment is where intelligence lives.
The Disinfection By-Product Case
No application of pre-treatment intelligence is more consequential — or more relevant to South Africa’s current regulatory trajectory — than the management of disinfection by-products (DBPs).
DBPs are not contaminants that arrive in the source water. They are formed during treatment itself, specifically when chlorine-based disinfectants react with natural organic matter already present in the raw water. The dominant regulated classes — trihalomethanes (THMs) and haloacetic acids (HAAs) — are therefore a function of two variables: disinfectant dose and precursor organic load. Treatment operators can control the first variable directly. The second variable is determined entirely by what was in the water before treatment began.
Regulatory Context — South Africa
SANS 241:2015 (Edition 6) sets THM limits at 200 µg/L aggregate for Class I water. The forthcoming SANS 241 Edition 7 is expected to introduce more granular DBP provisions and align more closely with WHO Guidelines for Drinking-water Quality (5th Ed., 2022) and USEPA Stage 2 Disinfectants and Disinfection By-Products Rule frameworks. Under either edition, a utility that does not routinely characterise its influent NOM loading has no early warning of elevated DBP formation risk — it discovers the problem in the effluent, after the exposure has already occurred.
This is the intelligence gap that pre-treatment monitoring closes. A water system with continuous influent NOM data — expressed as dissolved organic carbon (DOC), UV₂₅₄ absorbance, or specific ultraviolet absorbance (SUVA) — can model its DBP formation potential before adding chlorine. It can adjust coagulation to reduce the precursor load, modify disinfection contact time, or flag elevated-risk periods to its distribution network operators. None of this is possible if the first measurement is taken after the chlorine has already been added.
Industrial Pretreatment: The Upstream Attribution Problem
The Cincinnati Metropolitan Sewer District’s (MSD) experience with KETOS SHIELD continuous monitoring illustrates a second, less discussed dimension of pre-treatment intelligence: attributing water quality conditions to their actual sources before biological transformation obscures the origin.
MSD deployed continuous monitoring at both the raw influent and post-treatment effluent of their Little Miami wastewater treatment facility. What they found when they examined the nitrogen data across both points was counterintuitive: effluent nitrate was consistently higher than influent nitrate. The explanation is biological. The plant’s nitrification process was oxidising incoming ammonia into nitrate — which is precisely what a functioning biological treatment system does. The effluent nitrate elevation was evidence of healthy plant operation, not contamination.
“Comparing influent nitrate to effluent nitrate cannot tell you how much nitrate any specific industrial user is contributing, because the treatment process is adding nitrate by converting ammonia. The right place to characterise industrial nitrogen discharge is at the influent, as close to the industrial user’s connection point as possible, before biological transformation occurs.”— KETOS / Cincinnati MSD Case Study, 2024
The implication for South African water service authorities is direct. Municipal systems receiving mixed domestic and industrial influent — particularly those in industrial corridors around Johannesburg, eThekwini, and Nelson Mandela Bay — cannot reliably attribute effluent quality problems to upstream industrial dischargers on the basis of effluent data alone. The treatment process changes the chemistry. Pre-treatment monitoring at or near industrial connection points is the only analytical position that preserves the signal before the biology converts it into something else.
Continuous vs. Periodic: The Architecture of Insight
The shift from periodic grab sampling to continuous pre-treatment monitoring is not simply a matter of frequency. It changes what kind of knowledge is possible.
| Dimension | Periodic Grab Sampling | Continuous Pre-Treatment Monitoring |
|---|---|---|
| Temporal coverage | Point-in-time snapshots; monthly or quarterly | Full time-series; every 30–60 minutes, 24/7 |
| Event detection | Misses discharge events between sampling windows | Captures off-hours batches and transient spikes |
| Process insight | Lagging indicator; describes the past | Leading indicator; enables proactive dosing adjustment |
| DBP risk | NOM loading inferred retrospectively from THM results | NOM load measured upstream of disinfection; risk modelled in advance |
| Evidentiary value | Limited; single data point per sampling event | Timestamped, continuous record; defensible for billing or enforcement |
| Operator burden | Labour-intensive sampling coordination | Autonomous data collection; operator time spent on analysis |
The table above reflects a structural asymmetry: periodic sampling produces compliance records. Continuous pre-treatment monitoring produces operational intelligence. Both are necessary. Only one is sufficient for managing a water system in conditions of variability — which is to say, in real conditions.
What This Means for South African Water Systems
South Africa operates under a water quality regulatory framework anchored to SANS 241:2015, with 62 parameters across physical, chemical, microbiological, and radiological categories. The framework is rigorous at the point of supply — the tap. What it does not mandate, at the level of most municipal and small-system operations, is systematic influent characterisation upstream of treatment.
This gap has practical consequences that are compounding. South Africa’s infrastructure challenges — ageing water treatment works, underfunded water service authorities, persistent non-revenue water losses, and catchment degradation — mean that influent quality at many treatment plants is more variable today than it was a decade ago. Treating that variable influent with fixed chemical dosing protocols calibrated to historical averages is precisely the condition under which DBP formation risk rises, microbial breakthrough events occur, and effluent quality becomes inconsistent in ways that are difficult to diagnose after the fact.
- Catchment-level land use changes — expanding informal settlements, agricultural intensification, and industrial growth — alter source water NOM character and nutrient loading year by year.
- Climate variability in the Western Cape and Eastern Cape drives extreme fluctuations in turbidity and microbial load during high-rainfall events, which overwhelm treatment systems dimensioned for average-year inputs.
- Small water service authorities often lack the laboratory capacity for routine influent characterisation, creating a systemic intelligence gap at exactly the level of the system where data would be most consequential.
- Community-scale water treatment systems — point-of-use RO, ultrafiltration, and bioremediation units — operate with no pre-treatment monitoring at all, making membrane fouling events, chemical dosing errors, and microbial contamination effectively unpredictable.
Each of these conditions is a version of the same problem: a water system operating without the information it needs to make good decisions before the water changes. Pre-treatment monitoring does not solve these problems. It makes them visible early enough to solve.
Building the Intelligence Layer
The practical question for water systems at any scale — from a large metropolitan utility to a community-scale treatment unit — is how to begin building a pre-treatment intelligence capability without waiting for capital infrastructure programmes that may not arrive.
The answer is not always continuous automated monitoring. It is systematic, purposeful pre-treatment sampling that is understood as producing operational intelligence rather than compliance records. This means:
- Sampling at the right point: upstream of any chemical addition, upstream of any biological process, as close to the source as practically possible.
- Sampling at the right frequency: driven by process variability, not by regulatory schedule. Catchments with episodic industrial discharge or seasonal runoff patterns need more frequent influent characterisation than stable groundwater sources.
- Measuring the right parameters: NOM indicators (DOC, UV₂₅₄, SUVA), total Kjeldahl nitrogen (TKN), ammonia-N, turbidity, pH, alkalinity — the parameters that govern treatment chemistry and predict output quality.
- Recording data in a form that supports decision-making: timestamped, traceable, and linked to downstream process and effluent records so that the relationship between influent character and treatment outcome is visible over time.
- Using the data before acting: adjusting coagulant dose, modifying chlorine contact time, or flagging a potential DBP risk event — before the treatment process makes the condition invisible in the effluent.
This is the operational discipline that distinguishes a water system that knows what it is treating from one that discovers what it treated. The first system manages risk. The second one documents it.
Conclusion: The Question That Precedes Compliance
Water quality testing that begins at the point of compliance is testing that begins too late. By the time the water is measured at the tap, at the effluent discharge point, or at the distribution network boundary, the decisions that determined its quality have already been made — or failed to be made. The chemistry has run. The biology has transformed. The disinfection by-products, if they were going to form, have formed.
Pre-treatment data does not replace compliance monitoring. It precedes it, contextualises it, and makes it manageable. A laboratory result taken from raw influent before any treatment chemical has been added is not preliminary information. It is the most information-dense measurement available in the entire water system — a window into the source, the catchment, the upstream dischargers, the seasonal chemistry, and the process challenges that are coming.
Treating that window as an intelligence asset rather than a regulatory formality is the shift that distinguishes reactive water management from anticipatory water management. In a country managing the gap between SANS 241 compliance obligations and the physical realities of ageing infrastructure and variable source water, that distinction is not academic. It is operational.
References & Further Reading
- KETOS Inc. (2024). From Compliance Monitoring to Pretreatment Intelligence: How Cincinnati MSD Extended the Value of KETOS SHIELD. KETOS Case Study — Municipal Wastewater & Industrial Pretreatment. ketos.co.
- South African Bureau of Standards. (2015). SANS 241-1:2015 — Drinking Water: Microbiological, Physical, Aesthetic and Chemical Determinands. SABS, Pretoria.
- World Health Organization. (2022). Guidelines for Drinking-Water Quality: Fourth Edition Incorporating the First and Second Addenda. 5th Ed. WHO Press, Geneva. ISBN 978-92-4-004506-4.
- Richardson, S.D., & Kimura, S.Y. (2017). Emerging environmental contaminants: Challenges facing our next generation and potential engineering solutions. Environmental Technology & Innovation, 8, 40–56. https://doi.org/10.1016/j.eti.2017.04.001
- USEPA. (2006). National Primary Drinking Water Regulations: Stage 2 Disinfectants and Disinfection By-Products Rule. 40 CFR Parts 9, 141, and 142. United States Environmental Protection Agency.
- Chowdhury, S., Champagne, P., & McLellan, P.J. (2009). Models for predicting disinfection by-product (DBP) formation in drinking waters: A chronological review. Science of the Total Environment, 407(14), 4189–4206. https://doi.org/10.1016/j.scitotenv.2009.04.006
- Edzwald, J.K., & Tobiason, J.E. (1999). Enhanced coagulation: US requirements and a broader view. Water Science and Technology, 40(9), 63–70. https://doi.org/10.2166/wst.1999.0444
- Department of Water and Sanitation, Republic of South Africa. (2017). Blue Drop Regulations: Norms and Standards for Domestic Water and Sanitation Services. Government Gazette No. 41100. DWS, Pretoria.
- Van Zyl, P.G., Beukes, J.P., Du Toit, G., & Burger, J. (2014). Assessment of atmospheric trace metals in South Africa. Atmospheric Environment, 95, 25–39. https://doi.org/10.1016/j.atmosenv.2014.06.019
- Water Research Commission of South Africa. (2020). WRC Report No. TT 796/19: Natural Organic Matter in South African Source Waters: Character, Fate and Management. WRC, Pretoria. ISSN 1683-0917.
- Drinking Water Inspectorate, UK. (2023). Information Letter 03/2023: Guidance on Disinfection By-Products Monitoring and Risk Management. DWI, London.
- Muellner, M.G., Wagner, E.D., McCalla, K., Richardson, S.D., Woo, Y.T., & Plewa, M.J. (2007). Haloacetonitriles vs. regulated haloacetic acids: Are nitrogen-containing DBPs more toxic? Environmental Science & Technology, 41(2), 645–651. https://doi.org/10.1021/es0617441
The Chemistry Solutions Company (Pty) Ltd · WaterLab on Main
36 Clifton Road, Muizenberg, Cape Town · Partnered with SANS 241:2015 Accredited Water Quality Laboratory
Water Quality Intelligence Series · © 2026 TCSC · All rights reserved
0 Comments