QUAL2K and WASP are both free, both from the US EPA, and both widely accepted. They are not alternatives to each other in the way people often assume. QUAL2K is a focused tool for a river reach. WASP is a general water quality transport framework. Choosing the wrong one is not usually a question of accuracy, it is a question of whether the model can represent your water body at all.
QUAL2K, and what it is for
QUAL2K models water quality along a single, well mixed river reach, typically at steady state. It handles dissolved oxygen, biochemical oxygen demand, nutrients, algae, pH and temperature, with point and diffuse loads entering along the reach.
It is the right tool when the question is a classic river assimilative capacity problem. A discharge enters a river, and you need to know what happens to dissolved oxygen and nutrients downstream under low flow conditions. That is a very large share of real water quality work, and QUAL2K does it with far less setup than a general framework would demand.
Its limits follow directly from its design. One reach, one dimension, and a steady state assumption that is fine for a design low flow and wrong for a storm event. It does not do lakes, it does not do estuaries, and it does not do stratification.
WASP, and when you need it
WASP is a segmented transport framework rather than a river model. You define the water body as connected segments, and it solves advection, dispersion and a set of kinetic processes between them. That generality is the point.
Reach for WASP when:
- The water body is not a single reach. Lakes, reservoirs, estuaries, or a branching network that has to be represented as more than a line.
- Time-varying behaviour is the question. Storm loads, seasonal cycles, or a spill event. Anything where the answer is a time series rather than a profile.
- Sediment interaction matters. Contaminants exchanging with bed sediment, which is common in metals and persistent organics work.
- Toxics and bioaccumulation are in scope. WASP has modules for this that QUAL2K simply does not.
The cost of that generality is setup effort and a longer path to a defensible calibration. WASP will let you build something far too complicated for the data you have, and it is worth resisting.
Where CE-QUAL-W2 fits
CE-QUAL-W2 is the third name worth knowing here, and it occupies a specific niche: two dimensional and laterally averaged, which is exactly the shape of a stratified reservoir or a long narrow lake.
If your problem is thermal stratification in a reservoir, with a thermocline and different behaviour above and below it, neither QUAL2K nor WASP represents that naturally and CE-QUAL-W2 does. This is a case where the geometry of the water body decides the tool.
A decision rule
Draw the water body before choosing the model. If it is a line, and the question is a steady state profile downstream of a discharge, use QUAL2K and be finished quickly. If it is a shape, or the answer has to vary in time, or sediment and toxics are involved, use WASP. If it is deep, narrow and stratifies, use CE-QUAL-W2.
Licensing is not the deciding factor here, since the EPA models are freely available, though it is always worth confirming the current position with the developer. The deciding factor is representation, and the expensive mistake is starting in the wrong one and discovering three weeks in that your water body does not fit.
The part that is not about software
Water quality data is the hardest data in this field to trust. Probes foul, calibration drifts, and a fouled instrument and a genuine excursion look alike in a raw record. Any of these models will happily calibrate to a drifting sensor and give you a confident, wrong assimilative capacity.
Screening each instrument against its own history before it becomes a calibration target is not optional, and it is the step most often skipped. We treat it as part of the modelling work rather than as data cleaning, and it is a recurring theme in our water quality modelling.
We sell none of these tools. If you are scoping a discharge consent, a TMDL or a remediation target and are unsure which model the water body needs, that is a short conversation, and it is what our consulting and modelling services are organised around. Where the output has to become something a regulator or operator uses continuously, it becomes a decision support system instead of a study.