What NDMA is, how it forms, and why it is hard to remove.
Everything on this site comes back to one compound. This page explains it, topic by topic, with the source documents for each.
NDMA is not an exotic industrial pollutant that showed up in Wichita Falls by accident. It is a byproduct of the way most American water systems disinfect their water, it is measured in parts per trillion, and it is one of the few contaminants that reverse osmosis does not reliably take out. Each section below covers one piece of that, and lists the documents it rests on.
What NDMA is
N-nitrosodimethylamine (NDMA) is a member of a chemical family called nitrosamines. It is a very small, uncharged, water-soluble molecule, and those three properties explain most of its behavior in a water treatment plant.
Because it is small and uncharged, it slips through filtration steps that capture larger or charged compounds. Because it is water-soluble, it does not settle out or stick to surfaces. Because it absorbs ultraviolet light efficiently, it breaks apart under UV, which is the basis of the main treatment used against it.
NDMA first drew attention in drinking water in the late 1990s after contamination from rocket fuel manufacturing was found in California wells. Monitoring that followed led to a more important discovery: NDMA also forms during ordinary disinfection of drinking water and wastewater, which means it can appear in systems with no industrial source at all.
How NDMA forms
NDMA is a disinfection byproduct. It forms when certain nitrogen-containing compounds, called precursors, react with chloramine. Chloramination is used by many utilities because it produces fewer regulated byproducts than free chlorine and holds a more stable residual through the distribution system. For nitrosamines, the tradeoff runs the other way.
Precursors come from ordinary places. Wastewater is rich in them. So are some pharmaceuticals, some personal care products, and certain polymers used as coagulants in water treatment itself. A system does not need an industrial discharger to have NDMA precursors in its water.
"NDMA concentrations of 20 to 100 ng/L typically result from the chlorination of wastewater effluent."
Mitch et al. (2003), as cited in Plumlee et al. (2008)Two details matter for reading the Wichita Falls record. First, chloramination produces significantly more NDMA than free chlorination does. Second, formation is not instantaneous. It continues after the disinfectant is added, which is why where and when a sample is taken changes the number you get.
Wichita Falls chloraminates. That is not unusual and it is not improper. It is the reason NDMA is a reasonable thing to look for here.
How NDMA is measured
NDMA is measured by EPA Method 521, published in 2004 specifically for nitrosamines in drinking water. It is an established, commercially available test. Any accredited environmental laboratory can run it.
This matters because the request behind this site is not for novel research. It is for a standard analysis, using a federal method that has existed for more than twenty years, on water that is flowing right now.
Results come back in nanograms per liter (ng/L), also written as parts per trillion. That unit is easy to misread. One nanogram per liter is one part in a trillion, which is a million times smaller than one part per million. Published reference points for NDMA sit in the single digits and low tens of ng/L, so comparing an NDMA result to a ppm figure will always make it look like nothing.
The UCMR 2 minimum reporting level for NDMA was 2 ng/L. Wichita Falls results in 2009 ranged from 2.3 to 36.1 ng/L.
EPA UCMR 2, public water system TX2430001Health classification and standards
EPA classifies NDMA as a B2, probable human carcinogen, based on tumors induced at multiple sites in different mammal species. IARC classifies it in Group 2A, probably carcinogenic to humans, and the National Toxicology Program lists it as reasonably anticipated to be a human carcinogen. In animal studies the principal target organ is the liver, with lung and kidney tumors also reported.
There is no federal maximum contaminant level for NDMA in drinking water. That is the single most important fact on this page, and it cuts in both directions. It means no Wichita Falls result could ever be a violation. It also means there is no threshold anyone can point to in order to say the water is fine.
| Reference point | Value | Status |
|---|---|---|
| Federal maximum contaminant level | None | Not established |
| EPA tap water screening level | 0.11 ng/L | Screening value at a one in one million lifetime excess cancer risk |
| California Public Health Goal | 3 ng/L | Health based goal, not enforceable. A revised goal is in public review as of September 2026 |
| California notification level | 10 ng/L | Utility notifies its governing body |
| California response level | 300 ng/L | Removal from service recommended |
| California maximum contaminant level | None | Listed as pending |
California is used as a reference throughout this site because it has the most developed NDMA framework in the United States, not because its levels apply in Texas. They do not. Texas has no NDMA standard, and TCEQ has not adopted formal direct potable reuse regulations.
NDMA and reverse osmosis
Reverse osmosis is the centerpiece of almost every advanced water reuse system, and it is genuinely excellent at what it does. It removes salts, and rejection of most trace organic compounds of concern is high.
NDMA is the exception. The molecule is small enough and neutral enough to pass through RO membranes at rates that would be unacceptable for almost any other contaminant of concern.
"The percent removal of NDMA by RO at the IWPF was found to be fairly variable, with a range of 24 to 56 percent."
Plumlee et al. (2008), Water Research 42:347-355. Full scale Orange County Water District facility.The same study reports that thin film composite membranes typically achieve NDMA rejection of approximately 50 to 65 percent or less, and that rejection falls further with changing feed conditions and membrane fouling, which is why real operating plants perform below laboratory results. These figures appear in the body of the paper rather than its abstract.
There is a second point in that paper that is easy to miss. Microfiltration ahead of RO did not reduce NDMA at all. NDMA concentrations increased by 5 to 56 percent across the MF step, because the water was chloraminated before microfiltration and NDMA kept forming.
This is why potable reuse systems designed to control NDMA do not stop at RO. They add a UV step after it.
Wichita Falls documented both halves of this before the project began. A City wastewater emerging-contaminants assessment reported NDMA in the wastewater at up to 18 ng/L, and said in the same document that the removal rate across reverse osmosis had not been verified.
“Removal calculations … have not been projected because of removal rates from the RO have not been verified … the affects of the chloramines to the compounds hasn't been researched enough.”
City of Wichita Falls, Wastewater Emerging Contaminants Report (Appendix O), cover marked 2012-257. Spelling as in the original. The document states no sampling dates, sample counts, detection limits, laboratory or method.NDMA and UV treatment
Ultraviolet light breaks NDMA apart directly. UV is described in the literature as the most accepted treatment for NDMA, and it is what utilities install when NDMA is the problem they are solving.
But UV is not one thing. The same lamps can be run at very different intensities, and the dose determines what the system is actually treating. This is the distinction that matters most in the Wichita Falls record.
"It is estimated that the dose required for an order of magnitude reduction in NDMA is 10 times that for equivalent virus removal in a drinking water system."
Mitch et al. (2003), as cited in Plumlee et al. (2008)A UV system sized to inactivate pathogens is not sized to destroy NDMA. It is not a weaker version of the same barrier. At pathogen doses, NDMA destruction is incidental rather than designed.
Plumlee et al. measured 43 to 66 percent NDMA removal across a UV-AOP step at a full scale facility, operating at the high doses that advanced oxidation requires. That same paper notes that the hydrogen peroxide in UV-AOP is not expected to improve NDMA destruction on its own. It is the UV dose that does the work.
When you read the Wichita Falls UV records, the question to hold in mind is not whether UV was present. It is what dose it ran at, and what that dose was credited for.
What the Wichita Falls UV system was credited for →NDMA in potable reuse
Potable reuse concentrates the NDMA problem for a simple reason: wastewater is where the precursors are. Reported NDMA concentrations in raw sewage run in the tens to hundreds of nanograms per liter, and conventional secondary treatment removes a highly variable fraction of it.
The industry has known this for a long time, and the successful systems are built around it. EPA's own 2012 reuse guidelines describe NDMA in reclaimed water exceeding action levels even after RO, and note that several California utilities installed UV advanced oxidation specifically to treat RO permeate for NDMA.
The Orange County facility kept finished water NDMA below 10 ng/L in part through a disinfection choice: it used chlorine rather than chloramine after treatment, and its RO step removed most of the remaining precursors.
Plumlee et al. (2008), Section 3.3.3That is the comparison worth sitting with. A facility built to handle NDMA used a full advanced treatment train, then avoided chloramination at the end. Wichita Falls chloraminates its finished water.
Regulation has moved since 2015. California adopted direct potable reuse regulations in 2023 that set treatment and monitoring requirements for contaminants of emerging concern. Texas has not adopted formal DPR regulations, a point made in a 2026 peer reviewed paper on the Wichita Falls project.
NDMA in the distribution system
A treatment plant result is not the whole story. NDMA continues to form after water leaves the plant, as long as chloramine residual and precursors are both still present. That means concentrations can be higher at the far end of a distribution system than at the point where the water was treated.
Nitrosamine concentrations in a drinking water distribution system "increased with residence time, suggesting that these nitrosamines are formed as a result of disinfection."
Zhao et al. (2006), as cited in Plumlee et al. (2008)This has a practical consequence for any testing program. Where samples are taken changes what the results mean. A single sample at a treatment plant entry point does not describe what comes out of a tap several miles away.
The 2009 UCMR 2 results show this pattern in Wichita Falls itself. Results varied substantially between entry points and distribution sites on the same sampling date, and the highest single result of the year, 36.1 ng/L, came from an entry point in October.
See all sixteen 2009 results by location →How this applies here.
Everything above is general. These three pages apply it to the Wichita Falls record.
The 2009 results
Sixteen federal monitoring samples, fifteen detections, five years before potable reuse began.
View the results →What was tested
What the DPR project monitored, what it did not, and the documented sequence of recommendations and removals.
Read the record →The full timeline
Fifteen years of records, from the first federal detections through the current indirect potable reuse system.
View the timeline →Every document cited on this page is in the library.
Sources are grouped by tier so you can tell a federal guidance document from a peer reviewed study from a Wichita Falls record. How that sorting works, and where these documents came from, is set out in the methodology.