How-to Setup a Successful Compute

Study Dimensions

The three primary dimensions of an HEC-FDA Version 2.0 study are: project condition, analysis year, and impact area.

Project conditions, analysis years, and impact areas are not explicitly tracked by the software throughout all study data. We have left tracking the dimensions to which a given object corresponds (e.g., frequency function) up to the user. The user should track these dimensions using helpful naming conventions. Consider the following example name for a frequency function: "Analytical_WithoutProject_Year2023_ImpactArea1." The example name contains the type of frequency function, the project condition, the analysis year, and the impact area. 

Project Condition

A project condition reflects a given alternative plan, such as the no-action alternative (without-project condition), or one of the with-project alternative plans. If without-project condition damage is the sole interest, then only one project condition is required. If alternative plans are being evaluated, then one additional project condition is required to represent each additional alternative plan. 

A project condition is referred to as a plan in HEC-FDA Version 1.x. 

Analysis Year

An analysis year is the year that the study data represents. One or two analysis year can be specified (i.e., base year and most likely future year). If current expected annual damage and system performance metrics are the sole interest, only one analysis year is required. If average annual equivalent damage is also of interest (for net benefits and benefit-to-cost ratio) then two analysis years are required.

Impact Area

An impact area is a specific part of a study area to which the hydrologic, hydraulic, and economic relationships correspond. Review the Quick Start Guide's Impact Areas topic for more information.

An impact area is referred to as a damage reach in HEC-FDA Version 1.x.

Compute Configuration and Results

This page describes an example configuration for computing scenarios, alternatives, and alternative comparison reports. The examples in the illustrations assume that the hydrology and hydraulics reflect a naturally flowing river so that the discharge-frequency function can be represented using a Log Pearson Type III distribution and the stage-discharge function is relatively one-to-one. Departures from this example are described below. Each example represents a portion of the study data; the illustrated relationships apply generally. In the example, there are three impact areas (A, B, and C), two analysis years (base and future), a without-project alternative, and a with-project alternative. 

Scenarios 

A scenario consists of a set of hydrologic, hydraulic, and economic data for each impact area. The diagram below maps out the configuration of the base year and future year without-project condition scenarios. The appropriate set of stage-damage functions are selected for the scenario; then, for each impact area, the corresponding hydrologic and hydraulic functions are selected. Two more scenarios would need to be configured for this example: the base and future year with-project condition scenarios. The expected annual damage (EAD) distribution, EAD by damage category, and project performance statistics are the metrics available for each scenario compute. 

Alternatives and an Alternative Comparison Report

An alternative is configured by selecting the base and future year scenarios for a given project condition. In the example below, the base and future year without-project condition scenarios are selected for the without-project condition alternative. The same logic applies for the with-project condition alternative.  The results available for an alternative include the average annual equivalent (AAEQ) damage distribution and AAEQ damage by damage category. 

An alternative comparison report is a report that provides the comparison between the with- and without-project conditions. In the example below, there is one with-project condition alternative and the without-project condition alternative. The alternative comparison report is thus configured by selecting the two alternatives.  The results available for a given comparison include EAD reduced distribution, EAD reduced by damage category, AAEQ damage reduced distribution, and AAEQ damage reduced by damage category. 

Aggregated Stage-Damage Functions

The inputs required for an aggregated stage-damage compute are illustrated below. The wide cylinder represents a set of aggregated stage-damage functions for a given analysis year - project condition combination. Each parallelogram represents a data item or list of similar data items that must be identified as part of the configuration of a compute of a set of aggregated stage-damage functions. The compute requires a structure inventory that has been associated with valid occupancy types, hydraulic modeling, and a stage-discharge and discharge-frequency function for each impact area. A stage-frequency function may substituted in place of the stage-discharge and discharge-frequency functions if warranted, pursuant to the advice in EM 1110-2-1619 (i.e. stage-frequency should not be the default selection). A unique aggregated stage-damage function is computed for each impact area, damage category, asset category combination. 

Convergence Criteria

Convergence criteria by which all of the above mentioned results are computed can be configured by navigating to the Properties menu under File. Minimum and maximum iterations can be reduced for speedy preliminary results. The following default values should be used to compute results for reporting in decision documents and other related documentation: 

Impact Area Scenario Simulation (EAD and Performance Compute)

• Confidence limit at which to test convergence of the distribution: 0.95
• Tolerance of error: 0.01
• Minimum iterations: 10,000
• Maximum iterations: 500,000

Stage-Aggregated Damage with Uncertainty Computation

• Confidence limit at which to test convergence of the distribution: 0.95
• Tolerance of error: 0.01
• Minimum iterations: 500
• Maximum iterations: 5,000