FLOOD01888 - Laserfiche WebLink

Home Browse Search

EM 1110-2.1416 15 Oct 93 respunsibility of the hydraulic engineer to ensure that the level of detail is not too little nor too much for the stage of the study. (1) Although absolutes cannot be given regarding the level of detail for specific studies, Table 3-2 gives some representative guidance. In general, gradually varied steady flow is appropriate for most feasibility report analyses. Exceptions include those projects that obviously have an extensive effect on sediment regime (major channelization or reservoirs) that require movable buundary analysis in the feasibility phase, or those pro- jects that may significantly change velocity patterns or cause rapid changes in stage (locks and dams, power plant operations, etc.). Movable bed models and unsteady or multidimensional models are often utilized in the design stage, often after a data collection program has been in place 10 obtain the necessary data with which 10 calibrate and verify these more complex models. b. Data availability. While the first consideration should be study stage and level of detail required, the amount of available data also plays a part in the model selection. Gradually varied steady flow models can be calibrated with only highwater marks whereas movable boundary and unsteady or multidimensional models may require data from the entire hydrograph to calibrate. These models also require more hydraulic engineer skill and computer resources than gradually varied steady flow models. The necessity of using more sophisticated mod- els will usually become apparent in the planning process. Occasionally, higher level models must be used in the survey report stage, even without adequate calibration data. While the level of reliability may suffer due to limited or no calibration data, a skilled and experienced hydraulic engineer should be able to utilize such models 10 evaluate changes or differences due 10 a project, even though absolute with or without project values are ques- tionable. If accuracy is critical to the results of the feasi- bility report, a data collectinn program must be budgeted and planned for during the reporting process. c. Accuracy considerations. The term "accuracy" is rather nebulous when applied to hydrologic engineering. Physical and numerical models can yield information with a high level of precision, but with accuracy limited by the input data. The field data used to develop, cali- brate, verify, and operate models often vary :l:1O percent, or more, from the actual values. (1) The best evidence of the accuracy of the results is the skill and experience of the hydraulic engineer 3-14 performing the analysis. Rather than specifying a numer- ical range, an appropriate reply 10 an accuracy question might be: "Because the model has adequately repro- duced known events, the results for other, hypothetical, events are deemed 10 be representative of what would occur and results can be used with a reasonable level of confidence, provided that the same physical processes dominate in both known and hypothetical events." Implied in the foregoing is the use of sensitivity tests 10 evaluate the influence of key variables (Iike.n values) on design profiles to jodge the sensitivity of project econom- ics 10 those profiles. e (2) Determination of existing condition profiles requires the most care in the feasibility stage, as these profiles are key in the evaluation of existing potential damages, and flood hazard. Design studies require more accuracy for designing hydraulic components than neces- sary in the feasibility stage. . , . , d. Modeling requirements (time. experience, and computer resources). Modeling requirements vary with the reporting stage. In general, the more sophisticated the model required, the more time and cost is involved and the more limited is the pool of experienced engineers from which 10 draw. Only one or two experienced hydraulic engineers (at most) are usually available in any office 10 perform a hydraulic study requiring a multi- dimensional or movable boundary model. Other hydrau- lic engineers can encounter considerable start-up time and cost due 10 their inexperience with these techniques. e e. Hydraulic considerations. Computation of flow characteristics in natural channels can be a complicated and diffICult task. Many design failures and maintenance problems have resulted from the application of inade- quate or inappropriate analytical methods for the problem being considered. It is essential, therefore, to choose, develop, and calibrate the proper analytical method or modeling approach from the very beginning of a river hydraulics study. Much of the success of a project eval- uation lies in the ability 10 properly formulate the hydrau- lic studies as one of the ftrst tasks performed by the study team. The type of analysis needs 10 be accurately defined prior to selecting the model so that the study objectives dictate the model usage and not the other way around. , . . (1) As overviewed in Chapter 2, the classification and state of flow should be estimated as best as possible as an aid in selection of an analytical tool. Consider- ations are: e