HEC-RAS HYDRAULIC MODELING

REVIEW CHECKLIST
Gary W. Brunner and Cameron T. Ackerman

SCHEMATIC
Evaluate the model schematic to identify the complexity of the river system. This will let
you know what you are up against, as far as level of detail of the model, and amount of
time required to perform a review. You can also go to the HEC-RAS “File” menu and select
“Project Summary” to get a listing about the complexity of the model (i.e. number of rivers,
reaches, cross sections, bridges, culverts, inline structures, lateral structures, storage areas,
etc…). Here are some things to look for when reviewing the model schematic.

 How many river and reaches are there?

 How is the river system network connected? Is it a single reach system; dendritic
river system; does the system have looped reaches; is the system a very complex
network of interconnected channels?
 Are there lateral structures, storage areas, or split flow locations?
 How many hydraulic structures are there (bridges, culverts, inline structures, etc…)?
 How many boundary condition locations are there in the flow data file?

PROFILE
The slope of the river system should be evaluated very early in the review. Steep streams
are in general harder to model, especially for unsteady flow, as the water surface, flow rate,
and velocities change very quickly over short distances. Steep streams tend to have much
greater variability for cross section shapes, channel slopes, roughness, etc… This requires
more attention to changes in topography, land use, and channel roughness. Some things to
check very early in the review are:

 What are the slopes of the river reaches/land surface?

 Are there any significant changes in slope?
 Are there any significant drops in elevation between two cross sections?
 What is the slope of the computed water surface?
 What is the slope of the computed energy grade line?
 Where are critical/rapidly varied flow sections of the river?
 How much do the roughness coefficients vary?
 Did they use higher roughness values for the steeper portions of the river
 Is there a steady-flow plan/run to evaluate the model?

Steep slopes, large changes in slope, significant drops in the bed, dramatic channel
constrictions and expansions, and major changes in roughness are all areas that you should

HEC-RAS Modeling Checklist Page 1

pay close attention to. Make sure that the model represents what is physically there with
realistic terrain and model coefficients.

CROSS SECTION (XS) GEOMETRY

The next step is to make a thorough review of the geometric data in the model, which is
used to describe the channel, floodplain, and all of the hydraulic structures. Below is a list
of things to think about and check in the model:

 Are the Cross Sections reasonable? Do they accurately depict both the floodplain
and the low flow channel?
 Are the cross sections reach lengths appropriate for the size and slope of the river?
Remember, large flat sloping streams can have cross sections quite far apart.
However, small steep streams need short cross section spacing in order to account
for the rapid changes that occur in water surface, velocity, area, etc…
 Where did the data come from? In other words what was used to derive the cross
sections: surveyed sections, Lidar data, areal spot elevations, USGS DEM’s, a
combination of data sources, etc…
 Do the cross sections gradually transition or are there abrupt changes in geometry?
If there are abrupt changes, should additional cross sections be added to make the
transition more gradual between any two cross sections?
 Are any interpolated cross sections used in the model? This is ok for numerical
purposes, but are the interpolated cross sections appropriate? Or, should they have
gone back to the terrain model and cut more cross sections from the base data?
 Is there a low flow portion of the channel? Lidar Data, areal spot data, and USGS
DEM data generally does not include the low flow channel. Did they incorporate a
low flow channel? Where did they get the data? What assumptions were made? Did
they just cut a channel using the channel modification editor in HEC-RAS? If so,
what data did they use to support the development of the low flow channel.
 Were pilot channels employed? If so, what did they use for the width, depth, slope.
In general a pilot channel is just to help model stability, and it should not be used as
the main low flow channel.
 Are the main channel bank stations located appropriately (if the vertical portion of
cross section is imparting friction on flow in the center of the channel, bank stations
should include that part of the cross section as part of the main channel)?
 Are Manning’s n values appropriate? Are they higher in steeper reaches? Are they
reasonable for the stream being modeled? Do they accurately reflect the vegetation
and land use in the overbank areas?
 Was the use of horizontally varying Manning’s n values done appropriately for the
channel and floodplain?
 Did they use any flow versus roughness factors to account for changes in Manning’s
n value with flow/stage? Not all streams require this, but many large rivers do.
 Did they use Manning’s n values to model buildings in the flood plains? If so, are the
values they selected reasonable for the density and size of the structures?

HEC-RAS Modeling Checklist Page 2

  Are ineffective flow areas appropriately used to identify non-conveying portions of
the cross section and account for floodplain storage? Are ineffective areas turning
off together around bridges?
 Are blocked obstructions used appropriately?
 Is the Levee option used appropriately? If a levee is overtopped, is the levee option
turning off consistently? If there is a very large area/conveyance behind the levee,
which suddenly gets turned on, is this appropriate? Or should the levee be modeled
with a lateral structure, and the area behind the levee modeled with storage areas of
another reach.
 If this is an Unsteady flow model, do the pre-processed Hydraulic Tables for the
cross sections and the Bridges/culverts look reasonable?
 Were there any unique things in the geometry, such as debris blockages at bridges,
or ice options turned on? If so, the assumptions that went into modeling these
unique occurrences must be documented as to how they came up with the data and
values used to describe these features.

JUNCTIONS
 Are the junction lengths reasonable (are bounding sections too far apart)? If the
cross sections bounding a junction are too far apart, this can cause model stability
problems for unsteady flow modeling. Also, are all of the invert elevations of the
cross sections bounding a junction at about the same elevation? If not, at low flow
the model may experience stability problems.
 What method is used to compute the water surface profile across the junction? For
steady flow the default is the energy method, but the momentum method can be
used for high velocity junctions. For unsteady flow the default is the assumption that
the water surface is the same at all the cross sections bounding the junction. This
can lead to problems in steep reaches, or where the cross sections are spaced far
apart. A new option was added to HEC-RAS for unsteady flow, such that an energy
balanced is performed between the cross sections bounding the junction. This option
should be used when you know the water surface should vary with some significance
across the junctions.

BRIDGES/CULVERTS
Bridges and culverts can be difficult to model. This is especially true for structures that
cause a significant amount of contraction and expansion of the flow in order to pass through
the structure. Here are some things to check at Bridge and culvert crossings in the model:

 Are the cross sections laid out appropriately for the bridge/culvert?
 Are there any “complicated” bridges? i.e. complicated bridge openings, or multiple
opening bridge crossings, etc…
 Are the selected modeling methods and coefficients appropriate for the
bridge/culverts being modeled? Weir overflow for a bridge is normally set to 2.6, but
could range from 2.0 to 3.2 depending on the situation.
 Are there any bridge piers? If so, are the coefficients correct for the momentum and
Yarnell equations?

HEC-RAS Modeling Checklist Page 3

  Are the steady-flow expansion/contraction coefficients appropriate for the type of
constriction?
 Are ineffective areas appropriately set, both in station and in elevation? For high
events, are the upstream and downstream ineffective flow areas turning off
consistently together?
 Are Manning’s n values adjusted to reduce flow in the overbank areas just upstream
and downstream of the structure? This is a common technique in HEC-RAS to get a
consistent flow distribution upstream, through, and downstream of the
bridge/culvert.
 For unsteady flow modeling, are the pre-processed Hydraulic Tables reasonable for
the bridge/culvert crossings?

INLINE STRUCTURES
Inline structures can be used to model all sorts of hydraulic controls that cross the river
perpendicular to the flow. They can be used to model a dam, a simple weir, a drop
structure, or even a natural water fall.

 Is the top profile of the structure accurate? This is used to model any flow over the
structure.
 Does the inline structure have low flow openings (gates) and an overflow spillway?
 Are the coefficients appropriate for all of the hydraulic openings?
 Are the gate settings appropriate?
 Are the bounding cross sections appropriate? Does the upstream cross section
contain the full-flow cross section and have an invert that goes all the way to the
bottom of the inline structure?
 Does the structure have breach parameters? Are they reasonable?
o What is the final breach bottom elevation?
o Will the resulting invert be above, equal too, or below the cross section
invert?
o What is the breach Width, and is it reasonable for this type of structure?
o Are the side slopes within a reasonable range for a breach (0.1 to 1.0 H/V)
o What is the formation time of the breach (0.25 – 4hrs earthen, 0.1-0.5hrs
concrete). Some Dams and levees may be outside of this range based on
special circumstances.
o What is the breach progression method?

LATERAL STRUCTURES
Lateral structures can be used to model levees, constructed lateral structures, or even just
some high ground that the water may flow over. Lateral structures can often take a lot of
water out of the main river, so they need to be modeled accurately.

HEC-RAS Modeling Checklist Page 4

  What is the lateral structure connected to? (storage areas, or another reach)
 Is the weir profile (station –elevation points to describe the weir), length, and
overflow coefficients reasonable?
o Is the weir at constant elevation, sloped, or is it made up of many points to
describe the profile?
o Is the weir length long? Should it be broken in multiple, shorter structures?
 Is it properly connected to tail water cross sections of another reach? Or connected
to a storage area?
 Are there any gates, culverts, or rating curves used as part of the lateral structure?
If so, are the physical descriptions and hydraulic coefficients used appropriate?
 Does the structure have breach parameters? Are they reasonable?
o What is the final breach bottom elevation?
o Will the resulting invert be above, equal too, or below the cross section invert,
or the flood plain elevation? Is this an appropriate elevation for the bottom of
the breach?
o What is the breach Width, and is it reasonable for this type of structure?
o Size slopes (0.1 to 1.0 H/V)
o What is the formation time of the breach (0.25 –4 hrs earthen, 0.1-0.5hrs
concrete). Some Dams and levees may be outside of this range based on
special circumstances.
o What is the breach progression method?

STORAGE AREAS
Storage areas are used to model areas of ponded water. The can also be used as a set of
interconnected storage areas to model a more complex area, such as the interior area
behind a levee.

 Is the elevation volume curve reasonable?

 Is the curve smooth at the bottom of the curve? (i.e. are there enough points to
accurately describe the curve, especially at the lower elevations within the storage
area).
 What type of structures are being used to connect storage areas?

STORAGE AREA CONNECTIONS

Storage area connections are used to model hydraulic controls between two storage areas
(ponding areas). The can be simple overflow weirs, which would be common for modeling
flow going over a roadway, or they can be more complex with gates, culverts, and rating
curves.

 Are the weir profile, length, and coefficients reasonable for the storage area
connection?
o Is the weir at constant elevation, sloped, or is it a multi point line
representing the top profile of the structure?

HEC-RAS Modeling Checklist Page 5

  Does the storage area connection contain gates, culverts, flap gates, or rating
curves? If so are these hydraulic openings modeled accurately and are their
hydraulic coefficients appropriate?

PUMP STATIONS
Pump stations can be used to model a group of pumps to transfer (pump) flow from one
location to another. They can be hooked up from one cross section to another, from a cross
section to a storage area, or between two storage areas.

 How many pumps are in the pump station?

 Are they being modeled as identical pumps in one group, or as separate groups?
 Is the pump data accurate?
 Are the pump on and off elevations appropriate?
 Are they using Pump Override rules? If so, do the rules make sense? And are the
rules doing what the user intended them to do?

FLOW DATA AND BOUNDARY CONDITIONS

 What is the peak flow for the event being modeled? Do the cross sections and the
cross section property tables (for unsteady flow) go up high enough to completely
contain this event?
 What is the base flow for the event being modeled? Is this base flow reasonable? Is
it too high? Is it too low for the purpose of the model?
 What is the time of rise for the inflow or breach hydrographs? Time of rise can be
useful in figuring out an appropriate computational time interval for routing the
event.
 What is the downstream boundary condition(s) and is it reasonable?
 If the downstream boundary condition is not a know stage hydrograph (i.e. rating
curve, normal depth) is it far enough away from the study area, such that errors in
computed stage from this boundary condition do not affect the results in the study
area?
 What are the initial conditions settings?
o Gate openings
o Internal WSE for upstream XS
o Flows (Flows should be high enough to get the model running but <10% of
the peak flow so as to not affect the final results)
o Initial water surface elevations for storage areas connected to the river need
to be set at the same elevation as the initial stage in the river (if the
connection is such that flow can go back and forth at the initial condition)
 What types of boundary conditions are being used to model gates? Are these set
appropriately?
 Do they have all of the flow boundary conditions that are necessary to cover the
entire hydrology of the event? In other words are they accounting for all tributary
inflows? Are they also accounting for local runoff flows along the river?

HEC-RAS Modeling Checklist Page 6

  Are they using any of the special controls such as the Navigation Dam Operations
boundary condition, or the Rules editor for controlling hydraulic structures
operations? If so, is the data or rules they have entered correct and appropriate?

COMPUTATIONAL TIME STEP

The computational time step is one of the most crucial parameters to set when running an
unsteady flow model. Below are a few things you should consider in order to figure out if
the computational time step is appropriate.

 What is the average cross section spacing (DX)? (Min/Max)

 What is the average velocity of the flood wave at its highest velocity point (Vx)?
This will usually be on the rising side of the hydrograph just before the peak flow.
 You can use the Courant condition formula to check for an appropriate time step.
The basic formula is DT< DX/Vw Where Vw is the flood wave speed. Vw can be
approximated as Vw = 1.5 Vx.
 What is the Computational Time Step (DT)?
 Are there any dam or levee breaches occurring in the model? This will require small
time steps in order to model these breaches accurately.
 Are any lateral structures/levees overtopped during the event? This may also
require the need for a smaller time step.
 Are any gates opening or closing very quickly? This too would require the use of a
smaller time step in order to model this accurately.

CALCULATION OPTIONS AND TOLERANCES

 What value of Theta was used? Is this a tidally influence system (Theta=0.6) or a
dam breach (Theta = 0.8-1.0), or just a natural rainfall runoff type of event (Theta
0.8 to 1.0). Wave propagation, from a tidal boundary, or gates opening and closing,
will require theta to be set as close to 0.6 as possible, in order to truly track the
propagation of a wave with any accuracy.
 What are the Water surface elevation solution tolerances (default = 0.02)? If the
user raised this value, why? In general they should not be raised, but can be
lowered for more accuracy.
 What is the storage area tolerance set to (default 0.02)? Again, in general, this
should not be raised, but can be lowered for more accuracy.
 What are the number of iterations set to (default = 20)? If they are running the
model with no iterations, try turning iterations on and see if it changes their results
to any significance. Using more than 20 iterations is generally not necessary or
helpful, except under rare circumstances.
 Are they using the model warm up feature? If so, how many time steps are they
using to warm up the model? Is it reasonable? Model warm up is often necessary to
get a complex model started, so this is not unusual.
 Are they using time slicing? If so, what is the minimum time step set to? Is it
reasonable?

HEC-RAS Modeling Checklist Page 7

  Are they using any of the flow stability or submergence decay exponent features to
help model stability? If so, will the model run without them? Sometimes these are a
necessary evil, and required to get the model to run in a stable manner.
 Is the Mixed Flow Regime Option being used? If so, are the Froude number
threshold and exponent (m) set correctly?

MODEL CALIBRATION AND OBSERVED\HISTORIC DATA

Model calibration is an essential step in creating an accurate and reliable model. Pay close
attention to what calibration was done, and the results of the calibration. If no calibration
was performed, this should be a red flag on the model. Ask them why no calibration was
performed. Is there truly no observed data from historical events? There does not need to
be a gage to calibrate a model, historical high water marks can be use to calibrate a model.
A combination of both is ideal, but not always possible.

 Are there observed data? What is the observed data? Is it gauged stages and
flows? Is it high water marks? Where was this information obtained from and what
is the quality of the information?
 Was the model calibrated to observed data? If so, how many events? Or was a
period of record calibration run performed? How many years?
 How does the historic events compare to model results? In other words, is it a good
calibration? Could it be improved? Should roughness coefficients be varied with flow
and/or stage?
 Does the model perform well for high flows, medium flows, and low flows? This may
depend on the purpose of the model. If the model is only meant for flood studies, it
may not be important as to how well it does at low flow.
 Were any events held out from the calibration process in order to perform a
verification analysis? (this is running events that were not used to calibrate the
model in order to see how well it does). This can provide some insight on how well it
may perform for future event analysis.

UNIQUE ASPECTS OF THE STUDY AREA

Every watershed and river system is unique to some extent. Some systems may have
unique features that affect the river hydraulics analyses, such as: gate closer structures that
are only used during extreme events; levee or dike breaches that occurred during historic
events (this could greatly affect historic event calibration); special diversion structures;
etc… It is up to the modeler to know about unique things that happened during historic
events in order to have a good model calibration. It is also important to have knowledge of
unique features/structures in order to develop a good model. So as a reviewer, you may
need to do some investigation to obtain knowledge of unique events (for historic event
reproduction) and features of the system (in order to ensure the model is capable of
representing the true system.

HEC-RAS Modeling Checklist Page 8

HYDRAULIC UNCERTAINTY ANALYSIS
It is generally appropriate to perform some sort of parameter sensitivity analysis or a more
involved model parameter uncertainty analysis. If either were performed, the following
should be checked?

 What variables were used in the model sensitivity analysis or the more detailed
uncertainty analysis?
 Were the appropriate ranges of values used for the model sensitivity analysis?
 Were the results consistent with the directions of the parameter changes?
 Were any unique things like bridge debris or ice modeled, and therefore used in the
uncertainty analysis? How were ranges for these values arrived at?

HEC-RAS Modeling Checklist Page 9