Using QUALTX and DAFLOW/BLTM to Simulate Water Quality in Texas Rivers

USGS, Water Resources Division
208 Carroll Bldg., 8600 LaSalle Rd.
Towson, MD 21286
Internet: jslizar@usgs.gov
Phone: (410) 512-4902
Fax: (410) 512-4810

Introduction

The gradual realization that control of nonpoint-source discharges will be required to meet the nation's goal of "fishable and swimmable" waters has caused a gradual increase in the amount of federal, state, and local regulation with regard to such sources. This has had important policy and scientific ramifications for the state of Texas and will likely cause significant changes in the way in which water,our most precious natural resource, will be assessed and managed in the future. Section 303(d) of the Clean Water Act (CWA) requires states to develop Total Maximum Daily Load (TMDL) allocations for every water quality-limited stream segment in the state. A TMDL involves the determination of the appropriate limits to the quantities and quality of flows that can be discharged into our nation's surface waters. A segment is a defined stretch of surface water (lake, river, or tributary) which must meet certain water quality standards (WQS). The USEPA definition of a water quality-limited segment is any waters where existing or proposed controls do not or are not expected to result in attainment of WQS. Texas has a slightly expanded definition of water quality-limited segments. A segment is classified as water quality-limited if:

• Stream monitoring data have shown significant violations of the water-quality standards established by the State of Texas, or
• Advanced waste water treatment for point source waste water discharges is required to meet water-quality standards or to protect existing conditions of exceptional water quality, or
• The segment is a domestic water supply reservoir.

To fulfill the requirements of CWA 303(d), assess water quality, and perform TMDL's, Texas water quality-limited segments are currently evaluated (monitored and mathematically modeled) solely in terms of the point-source loading they receive. There is a well-established monitoring program and associated modeling effort that is used to establish permit levels based upon steady-state low-flow conditions in each water quality-limited segment. The low-flow conditions are considered to be the most critical conditions that could occur in a segment. They are critical because of the potential for water quality standards to be violated. When these conditions exist, a considerable percentage of the river flow is comprised of wastewater discharge.

The state environmental regulatory agency, the Texas Natural Resource and Conservation Commission (TNRCC), uses a steady-state surface water-quality model called QUALTX, a modified version of USEPA model QUAL-2E to evaluate permit levels for continuous point source dischargers around the state. During a period in which TNRCC considers a water quality-limited segment to be experiencing low-flow conditions, TNRCC staff visits the river segment to characterize the channel geometry, measure flow and discharges, and take various water quality samples. These are called intensive surveys. The data gathered on such an intensive survey enables the QUALTX model to be set up and executed to simulate the effects on water quality of various potential permit levels for point source dischargers. Based on the results, TNRCC makes recommendations on how, or if, existing permit levels should be changed. The establishment of permit levels based upon a modeling analysis is called a waste load allocation (WLA).

Neither nonpoint sources nor groundwater iinflows are considered when industrial, commercial, and municipal discharge limits are set through the use of QUALTX. Changes made to calibrate the model, which often include the addition or subtraction of flows to maintain the flow balance, are attributed to nonpoint-sources. But these gains and losses have not been verified with a gain and loss study typically used for that type of determination. In addition, not all the river segments across the state have been intensively surveyed. In some river segments, the TNRCC has based stream standards and permit limits on various grab samples and conventional permit limits.

While this approach has been acceptable in the past, the USEPA has begun emphasizing the importance of regarding water quality problems on a watershed basis and considering impacts from nonpoint sources. In addition, industry has convinced the USEPA of the necessity to seek tradeoffs between permit levels for point sources and the implementation of nonpoint source controls. These items are now supposed to be considered within the TMDL framework. This is quite a change from the historical command-and-control methods of simply targeting point sources and either performing waste load allocations based upon point source discharges or directing industries to apply the best available technology (BAT). This shift requires rethinking the current monitoring and modeling that the state of Texas performs to meet requirements of Clean Water Act 303(d).

Several case studies have been used to demonstrate the techniques involved in performing this new type of TMDL analysis. The USEPA has offered numerous workshops on the topic. The analysis involves a simple screening procedure for a comparison of point and nonpoint loads, as well as a more detailed modeling study for each affected segment. There are a number of more widely used flow and transport models that are described by the USEPA for possible use in a Compendium of Watershed-Scale Models for TMDL Development.

There appears to be ample room for individual state creativity in terms of the monitoring and modeling tools that are employed. Although these workshops have been held and several case studies have been provided to the TNRCC, a lack of more specific guidance and financial assistance from the federal government has caused some confusion as to how TMDL's can routinely be performed in Texas.

Thus far, neither TNRCC nor the USEPA has performed a TMDL on any of the Texas water quality-limited segments. But in response to this same evolution in thinking about water pollution problems, the Texas Legislature passed Senate Bill 818 in May of 1991. One of the major objectives of Senate Bill 818 is to qualify and quantify the impacts of nonpoint pollution relative to the impacts from point source pollution in every river basin. The act has been renamed the Texas Clean Rivers Act. Under this act, TNRCC is required to report every biennium on the water quality in Texas. This responsibility was passed on to the respective river authorities throughout Texas, with TNRCC responsible for assessing the water quality in those river basins without river authorities.

While no water quality modeling effort is currently required of the Texas river authorities under the Texas Clean Rivers Program, it is likely that it will be a required tool in the future to assess and manage nonpoint source impacts. One of the most advantageous aspects of such a modeling tool is its ability to simulate different potential scenarios once it has been calibrated. This is needed both for the purposes of the state and the river authorities, and for all water quality managers. Appropriate management decisions between the use of best management practices (BMP's) to control runoff or more stringent discharge requirements for continuous dischargers will rely on the formulation of models.

Another reason for the development of a new modeling approach for the state of Texas is that the state issues permits to various governmental entities and industries for the discharge of stormwater. Currently, there are no deterministic means of setting discharge requirements for stormwater. The impacts from stormwater are controlled by setting allowable water quality constituent concentrations at a level that is technologically and economically feasible for the discharger. Stormwater dischargers are required to implement Best Management Practices (BMP's) to control the runoff quantity and quality of stormwater. Neither the USEPA nor TNRCC has published any regulatory or economic justification which provides information on the actual improvements in the receiving water quality that will result from BMP's.

A deterministic modeling approach is needed that will allow for nonsteady- state processes to be simulated enabling hypothetical and real events to be modeled. Although this type of modeling is more sophisticated and data requirements are generally higher, the technology is available to meet the task.

Wherever possible, a new modeling approach needs to build upon data collection that has already occurred throughout the state. It also must be relatively routine in order to apply it on a broad scale across Texas. There has been some consideration of switching from the use of QUALTX to the use of a nonsteady-state model. This action has been delayed due to other TNRCC concerns, limited funding and research capabilities, and numerous agency reorganizations.

Determination of new critical flow levels for different water-quality constituents must be undertaken. "Critical storms" must be defined. This determination will most likely need to take into account the level of disruption caused by the storm flows to a river channel's physical and biological nature (causing bottom sediments to be disturbed thereby releasing pollutants) and the impact on the receiving water of the runoff quality from the different land uses that are present where the storm flows occur. Water-quality constituents that have historically not been considered when performing a WLA, and that are largely a result of nonpoint source pollution, must be incorporated into models. These include suspended solids and toxic substances. Data are becoming available that would aid in the development of a new dynamic modeling approach for Texas water quality-limited segments. Over $600,000 was granted to the various river authorities to perform nonpoint pilot projects under the Texas Clean Rivers Program. These projects have focused largely on data collection. Results have not yet been incorporated into dynamic deterministic modeling efforts to assess nonpoint source pollution impacts and re- evaluate existing permit levels.

Computer tools are also being developed that will aid in the simulation of nonpoint source pollution impacts. One of the most significant will be the development of spatial databases, also called geographic information systems (GIS). Tools such as GIS/ArcInfo, developed in Redlands, California by the Environmental Systems Research Institute, and recommended by TNRCC for river authority use in pursuing the objectives of the Texas Clean Rivers Program, can aid in the evaluation of spatially-varying data and its impact on water quality. It can also be utilized as a pre- and post-processor for model data.

Objectives

To accomplish this overall objective, the following specific objectives were completed:

• Compiled and reformated available data from an intensive survey and a steady-state model application of QUALTX as input for a nonsteady-state flow and water-quality model (DAFLOW/BLTM).
• Duplicated the results of TNRCC's steady-state model with the nonsteady-state model in a portion of a river basin.
• Modified the water-quality kinetics equations, located in a subroutine of the nonsteady-state model, to more closely emulate the kinetics of the QUALTX model. Added additional kinetic equations for water quality constituents typically caused by nonpoint source pollution.
• In the same portion of the river basin, used the nonsteady-state model to simulate the effects of a hypothetical storm event on receiving water quality.
• Evaluated the potential of this model for use in assessing WLAs and TMDLs in Texas.

Scope

An intensive survey of Segment 0611, a water quality-limited segment in the Angelina and Neches River Basin, as well as the input and results of a QUALTX model application was used to formulate the input for a DAFLOW/BLTM model application. DAFLOW/BLTM model results were compared with the QUALTX results for the same steady-state conditions. As a result of this comparison, the water-quality constituent kinetics of the BLTM model were modified to more closely emulate the kinetics of the QUALTX subroutines. This procedure involved recoding the FORTRAN of a BLTM subroutine.

Phase II required the creation of a hypothetical storm event to be superimposed on the steady-state conditions in the same basin. A USGS software program was used to generate storm hydrographs for a subbasin of Segment 0611. Typical literature values for the runoff concentrations associated with the storm flows were used. The dynamic input to the DAFLOW/BLTM model was formulated from this information. Model results were presented in such a way as to illuminate the utility of this new model for capturing the impact of a storm event and showing that a storm event could cause water quality standards to be violated. Alternative means of generating real and hypothetical storm event data were explored.