What FEMA Accepts on a LOMR, and Where the Time Actually Goes

FEMA name and U.S. Department of Homeland Security seal on the exterior of a federal office building.

Summary: On a National Flood Insurance Program submittal, the schedule is set by review cycles and engine-version compliance, not by how fast the model solves. FEMA accepts named versions of programs such as HEC-RAS, and generally expects a Letter of Map Revision to run revised conditions on the effective study’s version.

Most riverine Letters of Map Revision run on HEC-RAS. What surprises engineers filing their first one is that the release doing the computation is usually not theirs to choose. FEMA accepts programs at named versions, and the version is part of what you submit.

That one constraint shapes a lot of project schedules, and it is not where most teams look first. Solve time is the part of a National Flood Insurance Program (NFIP) study we tend to measure, because it is the part with a progress bar. It is rarely the part that costs the most calendar time. The LOMR is the NFIP submittal most engineers meet, so it is the one worked through here.

What FEMA accepts, and what it does not

FEMA does not accept modelers, workstations, or user interfaces. It accepts the computer programs that perform the analysis, and it names them in specific versions. The same rule covers every NFIP submittal: a restudy, a physical map revision, a CLOMR, or a LOMR.

The rule is 44 CFR 65.6(a)(6). Any program behind a map revision must be:

  • Reviewed and accepted by a governmental agency responsible for the implementation of programs for flood control and/or the regulation of floodplain lands.
  • Well-documented, including source code and user manuals; and
  • Available to FEMA and all present and future parties impacted by flood insurance mapping developed or amended through the use of the program.

That is why FEMA publishes a list of accepted numerical models rather than approving brands. HEC-RAS, built by the U.S. Army Corps of Engineers Hydrologic Engineering Center, sits on the hydraulic list.

The version rule lives in FEMA’s MT-2 instructions. Form 2 covers riverine hydrology and hydraulics, and it carries one line that governs a surprising number of schedules: where the effective model is being updated to produce revised conditions, submitters should “run the revised conditions modeling using the same model version as the effective, unless the entire modeled reach is being revised.” The reason FEMA gives is to avoid discrepancies outside the revised reach.

A modeler should not automatically open an effective model in the newest HEC-RAS release and recompute it. For a partial-reach revision, FEMA generally expects the revised conditions to use the effective model’s version.

Different modeling approaches may be appropriate when the entire modeled reach is being revised or when FEMA’s guidance permits another approach. A project that cannot meet those limits files a CLOMR under 44 CFR 65.12 first, and the map changes later through a LOMR.

The newest engine may well be better. For that submittal, it is still the wrong one. Version control on a LOMR is compliance work, not housekeeping.

LOMR or CLOMR: Which letter are you filing?

Both letters follow the same version rule, but they sit at opposite ends of a project. FEMA defines a LOMR as a letter “officially revising the current NFIP map to show changes to floodplains, regulatory floodways, or flood elevations.” A CLOMR comments on whether a proposed project would meet minimum NFIP standards if built as proposed. The CLOMR comments before construction; the LOMR revises the map after.

What forces the conditional route is the encroachment rule. 44 CFR 60.3(d)(3) tells communities to prohibit development in an adopted regulatory floodway unless analysis shows it “would not result in any increase in flood levels.” Where no floodway has been designated, paragraph (c)(10) applies instead. A project that cannot meet the applicable criterion may need a CLOMR under 44 CFR 65.12 before the community permits it, with the map revised later through a LOMR.

ConditionNFIP criterion
Adopted regulatory floodwayProposed encroachment must not increase flood levels
AE/A1-30 without designated floodwayCumulative effect must not increase BFE > 1 ft
Increase beyond applicable criterionConditional map revision may be required

One practical detail is easy to miss: not every submittal goes to the same place. Some states and regional agencies review their own map revisions under partnership agreements with FEMA, among them North Carolina, Alabama, Illinois, and the Urban Drainage and Flood Control District in metro Denver. Check the routing list in the current MT-2 instructions before you assemble the package.

Where the months really go

A LOMR analysis may involve several model conditions, and MT-2 Form 2 defines the principal model types used to document the revision:

  • Duplicate effective: A copy of the analysis used in the effective Flood Insurance Study.
  • Corrected effective: Fixes errors or adds cross sections, with no physical change.
  • Existing or pre-project conditions: Changes in the floodplain since the effective study.
  • Revised or post-project conditions: The condition being submitted.

They all have to agree with each other and with the published study. 44 CFR 65.6(a)(2) requires revised and unrevised base flood elevations to “match within one-half foot where such transitions occur.”

FEMA’s guidance for flood risk analysis and mapping also provides an additional important modeling check. At the limits of revision, post-project BFEs should generally match pre-project BFEs within 0.1 ft, while the regulatory transition requirement is 0.5 ft against the effective FIS BFEs, subject to the applicable exceptions.

Then there is the review clock. FEMA’s published MT-2 sequence allows up to 90 days to issue review comments, 90 days for you to address them, and 90 days to issue the revision, followed by a 4.5-month appeal period. A clean submittal can still run most of the year. Even without substantial technical rework, the published review and appeal process can place the overall calendar duration on the order of several months.

PhaseWhat drives itWhat compresses it
Duplicate effectiveFinding an effective model and matching its outputRunning the same engine version the effective study used
Plan managementKeeping four condition models consistentOne project holding every plan, with shared geometry and flow data
Floodway determinationIterating encroachment stations to the allowable surchargeImmediate feedback on rise, width, and velocity
FEMA reviewThe published process requires up to 90 days each for:
  • FEMA’s review and initial action,
  • Addressing comments by the requester, and
  • FEMA to issue the determination
Submitting something that does not draw a second round of comments

The last row is the one worth staring at. Every avoidable warning is a possible review cycle, and a review cycle is measured in months, not hours.

Why a faster computer will not help

There is a persistent belief that HEC-RAS run times are a hardware problem. The HEC-RAS User’s Manual says otherwise. In testing on a computer with 8 logical processors, “the performance gains were maximized for models using 4 cores.” Small models got slower as cores were added, and hyper-threading can work against you too.

USACE’s 2016 guidance on optimizing a computer for HEC-RAS is blunter: “Processor speed is still paramount. Do NOT think you will have fast HEC-RAS model run times just because you have a computer with 16 cores or more.”

Time step and mesh resolution move the needle further than core count. But they trade against accuracy, and on a regulatory floodway determination that is not a trade worth making to save an hour.

Recent releases make the tuning easier. HEC-RAS 7.0 arrived on April 17, 2026, and the 7.0.1 bug-fix release followed on June 2. Version 7.0 added a Run Multiple Plans option that runs one plan at several processor counts and collects the times in one table, and it now tells performance cores apart from efficiency cores.

Forward compatibility is not regulatory interchangeability. A newer engine will open an older project. That does not make it the version Form 2 expects.

What actually shortens the work

None of that changes with the interface a modeler uses. GeoHECRAS runs the USACE HEC-RAS engine unmodified, at the version the user selects. It is an easy-to-use 2D/3D graphical user interface data wrapper to the US Army Corps of Engineers HEC-RAS. The engine result is the engine result. What changes is the work wrapped around the compute, and that is where the schedule on any NFIP submittal lives. Four things move it:

  • Engine version selection. You pick which HEC-RAS engine computes, from 7.0.1 back to 3.0.1, and you can lock it for imported models. Against Form 2’s same-version rule, which replaces a shelf of separate installations. To learn more, refer to this article in our knowledge base.
  • Automatic core selection. The software reads the processor configuration and picks the solver cores, favoring performance cores, with a manual override. As the Knowledge Base puts it, “Selecting ‘All Available’ solver cores does not always produce the fastest simulations.” To learn more, refer to this article in our knowledge base.
  • One project, every plan. All four condition models sit side by side, each with its own geometry, flow data, and terrain, instead of in four folders with hopeful filenames. Batch analysis recomputes the ones you select, unattended. To learn more, refer to this article in our knowledge base.
  • Floodway iteration. Encroachment stations compute while you watch the rise, the width, and the velocities. A copy step moves a Method 4 solution into explicit Method 1 stations, which some state reviewers require. To learn more, refer to this article in our knowledge base.

The savings customers measure land where the theory says they should: in model development, not in compute. Civil & Environmental Consultants reported cutting model development time by 50 percent, with model building and calibration falling from about half the project budget to a quarter.

None of it transfers engineering responsibility. Sections cut from terrain still need review where the terrain is wrong, roughness from land cover still needs field judgment, and every submittal still carries an engineer’s seal.

Five things to check before you submit

FEMA hands you its review checklist in advance, which makes the cheapest schedule win also the most overlooked. Its cHECk-RAS tool reads HEC-RAS data files and flags inconsistencies across five checks: Manning’s n and transition losses, cross sections, floodway, structures, and profiles. Run it before the package goes out.

  1. The engine version matches the effective study, and the report says so in writing rather than leaving a reviewer to infer it.
  2. The duplicate effective ties back, and the comparison is documented. Keep the three figures distinct.
    • FEMA requirement
      0.5 ft transition criterion under 44 CFR 65.6(a)(2), subject to applicable conditions/exceptions.
    • Modeling/practice check
      0.1 ft tie-in comparison may be used in FEMA guidance for post-project vs. pre-project limits.
    • State/local guidance
      Any additional 0.1/0.5-ft requirement from Illinois should be explicitly labeled as Illinois guidance, not FEMA policy.
  3. Revised and unrevised base flood elevations match within half a foot at the transitions, per 44 CFR 65.6(a)(2).
  4. cHECk-RAS runs clean, or every warning left has a written explanation a reviewer can accept.
  5. The package is going to the right office, according to the routing list in the current MT-2 instructions.

So, if the engine is fixed by regulation and the compute is a small share of the calendar, why do so many teams still measure a LOMR in solver minutes?

“Faster” on a map revision does not necessarily mean a shorter compute. It means fewer review cycles, fewer rebuilt models, fewer version surprises, and fewer hours spent carrying numbers between windows. The engine is fixed by regulation, exactly as it should be. Everything around it is ours to improve, and that is where the months are hiding.

Frequently Asked Questions

Generally, the same version as the effective model should be used. FEMA’s MT-2 Form 2 instructions direct submitters to run revised conditions on the effective model’s version unless the entire reach is being revised. Confirm it with the reviewing agency first.

FEMA describes a LOMR as a letter officially revising the current NFIP map to show changes to floodplains, regulatory floodways, or flood elevations. A CLOMR is a letter commenting on whether a proposed project, if built as proposed, would meet minimum NFIP standards. The CLOMR comments before construction; the LOMR revises the map afterward.

Not necessarily. In USACE testing, gains were maximized at four cores; small models can run slower as cores are added, and hyper-threading can reduce performance. Processor speed matters more than core count, and on a LOMR the compute is rarely the constraint anyway.