Develop Improved Guidance for Dam Break Forecasting

8/1/2008

Although dam failures are infrequent, the impacts can be catastrophic. The National Weather Service (NWS) must issue accurate and timely forecasts of flood waves resulting from dam failures to protect lives and property. Although already an important part of the NWS mission, accurate dam break forecasts will become increasingly important during the next 25 years. In 2020, approximately 80 percent of dams in the United States will be more than 50 years old (Canino and Taylor, 2004; Koehler, 2008)

Dam failures can be sudden events and the resulting flood waves travel rapidly downstream; therefore, forecasters need pre-calculated scenarios in many situations. However, it may not be possible to pre-calculate all important scenarios, and for this reason it will be beneficial to re-run models during events using real-time information (particularly with information about how quickly the dam is failing and the current water level). One of the biggest engineering challenges is predicting the time it will take a dam to fail, and this time to failure has a major influence on the flood characteristics at downstream points.

The dam break forecasting problem can be divided into three parts: (1) developing the data and models to accurately forecast dam breaks, (2) implementing these models in an effective manner (a.k.a. "Concept of Operations"), and (3) verifying forecasts and using this information to improve the models and forecast process.

The first part, the dam break modeling itself, can be further broken down into three steps. (1a) describing the structural dam failure mechanism (predicting time to failure and breach size (Wahl, 1998), (1b) propagating the released water to downstream points of interest, and (1c) mapping the flood extent and depths. The NWS has developed and fielded tools to execute each of these steps. The NWS Dam Break model and Flood Wave Dynamic Model (FLDWAV) (Fread, 1993) software contain a BREACH (Fread, 1985) subprogram that simulates piping and overtopping failures in earthen dams when users provide breach parameters. Flood propagation can be modeled using either FLDWAV or Simplified Dam Break (SMPDBK) program (Whetmore et al., 1991). The Dam Catalog Reviewer Estimator Tool (DamCREST) software provides forecasters a user interface to view and manage dam data and run SMPDBK within the Advanced Weather Interactive Processing System (AWIPS). In addition, Welch (2007) put together a simple graphical interface to compute breach parameters using empirical equations from the literature. This tool also has an option to perform simple dam break analysis using Rules of Thumb. The Geographic Information Systems (GIS)-based FLDVIEW software creates inundation maps using FLDWAV outputs (water elevations) and also assists the user in deriving cross-sections during the development of FLDWAV models.

Despite the existence of these tools, limited documentation for some components, lack of comprehensive training opportunities, changes in underling technologies (e.g., FLDVIEW relies on old Arcview 3.x GIS technology), or lack of functionality within AWIPS have limited effective implementation.

For this HSMB research project, The OHD/HSMB Hydraulics group proposes the following tasks to overcome these limitations and more effectively support dam break modeling at the River Forecast Centers (RFCs) and Weather Forecast Offices (WFOs):

1. Strengthen the HSMB expertise in dam break modeling to support RFCs and WFOs.

2. Prepare comprehensive and clear guidance on how to conduct dam break analyses – NWS Handbook for Dam Break Forecasting. This involves a number of subtasks including:

1. Recommend procedures to access and quality-control the most up-to-date dam information.
2. FLDWAV to Hydrologic Engineering Center - River Analysis System (HEC-RAS): Verify through literature review and experiments (as necessary) that HEC-RAS dam break is a viable replacement for FLDWAV dam break. We anticipate that it will be easier to set up new dam break models using HEC-RAS; however, we may need to maintain some level of capability with other dam break models in the NWS to work with existing Emergency Action Plans (EAPs). Ed Capone, Senior Hydrologist at the North East River Forecast Center (NERFC) notes that it is useful to be able to run existing EAP models that usually were prepared at a cost of thousands of dollars to make sure the NWS is talking the same language as Emergency Managers during events. Ed also notes that existing EAP dam failure scenarios may have been run using seven different models.
3. Document tradeoffs between using SMPDBK and more complex models such as HEC-RAS. One obvious limitation of SMPDBK is the inability to simulate river systems with multiple dams.
4. Identify a recommended GIS tool to cut cross sections and document easy-to-follow procedures. Defining reasonable cross-sections is important when using either the SMPDBK, FLDWAV, or theHEC-RAS dam break model. HEC-GeoRAS is one candidate tool. The RFCs need an easy to use and accessible tool.
5. Identify the most effective GIS tool to generate inundation maps and document easy-to-follow procedures. This task will be primarily addressed as part of a separate project on inundation mapping.

3. Contribute to dam break analysis training as appropriate.

4. Stay abreast of ongoing research by the US Army Corps of Engineers (USACE) and the US Bureau of Reclamation on dam failure mechanisms.

Aging Dams Threaten U.S. Water Supply Systems, AWRA Annual Meeting Conference Proceedings.

Fread, D.L., (1985). Breach: An Erosion Model for Eastern Dam Failures, NWS Hydrologic Research Lab, 34 pp.

Koehler, R., (2008). Dams and Dam Failure – Module 1: Terminology and Open Channel Hydraulics, http://www.meted.ucar.edu/resource_modlist.php.

Wahl, T. (1998). Prediction of embankment dam breach parameters- A literature review and needs assessment. DSO-98-004 Dam Safety Office, Department of Interior.

Welch, D (2007) Dam break Scenario Rules of Thumb used at LMRFC. Unpublished note.

Whetmore, J.N., Fread, D.L., Lewis, J.M., Wiele, S.M., (1991). The NWS Simplified Dam-break Flood Forecasting Model, 47 pp.