Loading...
Park, Peter-CS 960910lll 8616 NORTHWEST PLAZA DRIVE Halff A ociate mmm S S S Ill ~x ~-oo~ ENGINEERS · ARCHITECTS · SCIENTISTS PLANNERS · SURVFYORS September 10, 1996 AVO 15783 Mr. Kenneth M. Griffin, P.E. Assistant Ci~, Manager/Director of Public Works City of Coppell 255 Parkway Blvd. Coppell, Texas 75019 Fax: 214-393-0931 or 214-304-3570 Re: Flood plain review ofa CLOMR by others, for a 18.9 acre site on Sandy Lake Road, City. of Coppell, Texas. Dear Mr. Griffin: We have completed the review of a Request for a Conditional Letter of Map Revision (CLOMR) for a 18.9 acre site on Sandy Lake Road, City of Coppell, Texas, originally prepared by Nathan D. Maier Consulting Engineers (NDMCE) in July 1996. We have reviewed the CLOMR report for compliance with the Ci~r of Coppell Floodplain Management Ordinance No. 94639 approved on March 8, 1994 and, in the process of doing so, have reviewed the report for compliance with the FEMA requirements for a CLOMR submittal and good hydraulic engineering practices in general. In addition to this letter report, you will find included in this package eight (8) pages of notes and 2- 1 l"x17" maps prepared for this project by Halff Associates, Inc. The letter report discusses the main issues encountered in our review. The attached pages of notes cover additional items considered, but of lesser importance. The following paragraphs address deficiencies in the submittal based on the City of Coppell Floodplain Management Ordinance. Some of these matters are traditionally included in the CLOMR request, while others are often included in a separate hydraulic study for the City's review. The CLOMR mentions that a separate report is being prepared for a Corridor Development Certificate (CDC) submittal. 1. Impacts of the project on valle)' storage have not been addressed by the NDMCE report. Although FEMA does not require such information, valley storage preservation is a requirement of the Ci~' of Coppell Floodplain Management Ordinance and CDC guidelines. This appears to be the most serious deficiency in this submittal. DALLAS · FORT WORTH ° HOUSTON ° ARLINGTON · McALLEN ° CHICAGO TRANSPORTATION · WATER RESOURCES · LAND DEVELOPMENT ° MUNICIPAL · ENVIRONMENTAL. STRUCTURAL MECHANICAL · ELECTRICAL · SURVEYING · GEOGRAPHIC INFORMATION SYSTEMS ARCHITECTURE · LANDSCAPE ARCHITECTURE ° PLANNING Halff Associates ENGINEERS ARCttlTECTS SCIE. TISTS PLANNERS · SUR\"EYORS 2. The report deals only with CLOMR hydraulics (existing watershed conditions using FIS discharges). Because the report does not include a hydraulic study based on ultimate watershed conditions, those Floodplain Management Ordinance requirements that apply to fully developed watershed conditions are not met. Some of the missing information might be provided separately to the City in the form of a copy of the CDC submittal. The project's impacts on SPF elevations need to be determined as well. 3. Article 4, section C, also requests that all other necessaD' permits be obtained. That might include a Corps of Engineers 404 permit if applicable. Wetland considerations are not mentioned in the text of the CLOMR. The following paragraphs cover the items that we feel the Federal Emergency Management Agency (FEMA) might cormnent on in their response to this Request for a CLOMR. 1. The proposed project map (a grading plan) does not show the HEC-2 cross section locations nor the existing and proposed flood plain delineations and floodway. Only part of one cross section line (93810) is shown on the existing site plan. The proposed project map and the existing site plan are at a different scale so that they cannot be overlaid for comparison. This situation renders checking delineations and flood plain widths rather difficult. 2. The FIS model used by NDMCE for duplication might not be the effective one. The Sand)' Lake Road and McInnish Park Improvements (SLRI) request for a CLOMR was approved by FEMA on May 25, 1995. That model may be the effective FIS model for this site. The Sand), Lake Road bridge and roadway that soon will be constructed across the Elm Fork of the Trinity River are mentioned in the text of the NDMCE report but not included in the model that NDMCE used (the project site would have been shown almost entirely in an ineffective flow area if the Sandy Lake Road and McInnish Park Improvements CLOMR had been used). This issue should be confirmed by the applicant with FEMA. 3. Cross sections could have been added at several locations in the existing conditions model to represent the proposed fill with greater accuracy. The only cross section modified seems to merely touch the edge of the fill. The model does not include the peak of the fill. The fact that the next cross section upstream is 3760 feet away only worsens the situation. The same problem happens in the floodway model where no cross section reflects the true extend of the proposed changes. Those items might be serious enough for FEMA to ask for revisions to the models before they will agree to issue a CLOMR. · -- Halff Associates lll ENGINEERS · ARCHITECTS · SCIENTISTS PLANNERS · SURVEYORS The attached pages of notes cover all the questions and comments that were raised during om review of this CLOMR for compliance with the City of Coppell Floodplain Management Ordinance, for adherence with the FEMA requirements and general good hydraulic practices. They are included with this letter only to indicate the particular questions about this CLOMR answered in this review. The maps show the project site and HEC-2 cross sections against the Trinity River Feasibility Study Topographic maps - NDMCE prepared their own maps based on their own field sun'eys. As explained in the attached notes, although the NDMCE maps have a higher degree of accuracy (1 fi contours versus 2 fi) they do not show much of the surrounding area and show only one HEC-2 cross section. In summary, we recommend that the City of Coppell request additional information from the applicant, most importantly detailed analysis showing valley storage impacts. Correct hydraulic models and discharges should be used to determine impacts on flood levels, especially the SPF. Please call our offices at 214-739-0094, extension 226, if we can be of further help or if you have any questions. Sincerely, HALFF ASSOCIATES, INC. HALFF ASSOCIATES, INC. Sabine Borgnet-Harfis, E.I.T. Walter E. Skipwith, P.E. Project Manager and Vice-President Enclosures. File: 15783\word6'a-eport.doc Notes. Avo 15783 Date: August 29, 1996 Client: City of Coppell Subject: Flood plain review / NDMCE CLOMR Sandy Lake Road 18.9 Acre property 1) Miscellaneous. · SLRI refer to the improvements designed by HalffAssociates, Inc. in 1994 (avo 12687), hydraulics for Sandy Lake Road improvements and swale design in Mclnnish Park. · FYI only: Cross section labels mentioned below pertain to the FEMA models used by NDMCE and are different from the ones recently developed by the US Corps of Engineers for the Elm Fork of the Trinity River. · The SLRI were delineated against the Trinity River Feasibility Study aerial topography (TRFS in this text). 2) Exhibits included with this review. · One 1 lx17 map, scale 1"=200', with the general area as shown on the TRFS topographic maps. It includes FIS cross section lines. Those are the cross sections used in all the NDMCE models for this CLOMR submittal. · One 1 lx17 map, scale 1"=200', which shows both NDMCE cross section lines and SLRI revised cross section lines. This exhibit also shows the extent of effective flow per the proposed SLRI model. 3) Cit)' of Coppell Floodplain Management Ordinance requirements. · City permit forms for a flood plain development permit and drawings to scale. The NDMCE report we were given to review covers only the CLOMR submittal. · Inclusion in the report of technical data showing the effect of the proposed improvements on floodplains, floodways, flood elevations and flow velocities for the 2-year and the 100-year mean recurrence interval floods, and FIS discharges: the report includes hydraulic results generated by the Corps of Engineers HEC-2 computer program for thc 1 O-year, the 50-year, the 100-year and the 500-year mean recurrence interval floods, as required by FEMA for a CLOMR submittal. It does not include fully developed watershed hydraulics. · The lowest floor should be elevated to a minimum of two (2) feet above the FIS base flood elevation: the fill elevation varies between 447 and 449, while the FIS base flood elevation is 445.22. A future building finished floor elevation greater than 447.22 will meet the requirement. · The lowest floor should be elevated to a minimum of one (1) foot above the design base flood elevation: that criteria will be determined based on fully developed watershed hydraulics, not available with the CLOMR. · Inclusion in the report of delineations of existing and post-development floodplains. The delineations are included but on a separate map from the proposed grading plan; and at a different scale, which renders any comparison rather difficult. * The matter of valley storage has not been addressed by the NDMCE report. Although FEMA does not specifically require such information, there is no guarantee that the project as designed meets the City of Coppell requirement of: · a maximum 15 percent valley storage reduction for the 100-year design flood and a maximum 20 percent reduction for the Standard Project flood (drainage areas of less than 100 square miles), · a maximum 0 percent valley storage reduction for the 100-year design flood and a maximum 5 percent reduction for the Standard Project flood (drainage areas of more than 100 square miles) In this case, the project needs to meet the most stringent of the 2 requirements. · Demonstrate with technical data that the flood carrying capacity within the channel is maintained: that requirement has been met by the use of HEC-2. · The development must not increase the effective FIS base flood (existing condition watershed) by more than one (1) foot, when the combined effect of existing and proposed improvements are combined: that requirement is met. Revisions to the effective model for existing and proposed conditions combined result in a maximum increase of 0.00 foot per the HEC-2 natural condition (multi-profile) model outputs included in the CLOMR report, for the effective FIS base flood. Some confusion occurs when one reads the water surface elevation comparison table included ~vith the FEMA Form 2. The table quotes the floodway (encroached) models instead of the multi-profile models. There is a 0.01 foot discrepancy between the multi-profile and the floodway models. · The development must not increase the water surface elevation at all, if the site is located within the FIS regulator), floodway, when the combined effect of existing and proposed improvements are combined: that requirement is met. Please refer to 9) Floodway models, which explains that the floodway model never "sees" the proposed improvements. Also refer to the previous paragraph that deals with a 0.01 foot discrepancy that exists between the multi-profile and the floodway models. · The development must not increase the design flood (fully developed condition watershed) water surface elevation by more than zero (0) foot, w'hen the combined effect of existing and proposed improvements are combined: there is no guarantee that the requirement will be met until the fully developed watershed hydraulics can be examined. · Article 4, section C, also requests that all other necessary permits be obtained. That might include a Corps of Engineers 404 permit if applicable. Also, because the site is located along the Elm Fork of the Trinity River, a Corridor Development Certificate (CDC) is most likely necessary. The CDC is mentioned in the CLOMR but not the 404 permit. · Design discharges considered should include the mean annual flood (the 2-year) and the low-flow or base-flow discharge (dry weather flow or no less than 5 cubic feet per second per square mile): those are traditionally not included in a CLOMR submittal. · Improvements should not cause average channel velocities to exceed maximum permissible values for the 2-year mean recurrence interval flood: the average channel velocity for an)' computed flood at or upstream of the improvements remains below' 4.69 fps. · Alterations of the flood plain shall occur only if it can be shown that equal conveyance alterations can occur on both sides of the channel, while all the criteria of the ordinance are met: This requirement can be met by using encroachment method 4 of the HEC-2 program, but is not met in the regular execution of the program. It does not apply to letter of map revisions, the matter applies only to defining floodways for new areas. 4) Report and Maps · The report does not include photographs of the site. They are helpful to establish existing conditions but not expressly required by the City of Coppell or FEMA. · The report does not include a geotechnical report. The forms do not request it anywhere. · There are 3 maps included with this report, a location map bound with the book, a 24"x36" grading plan and a 24"x36" flood plain work map. · The FIRM delineations appear to be correctly transferred to flood plain, albeit roughly (see zone X as it touches the floodway). · Workmaps could use a legend - map line codes are clear, though. · Map shows only one HEC-2 cross section, 93810, it could have shown others downstream. Inconsequential. See attached plot by Halff Associates, Inc. with cross sections upstream and downstream of the site shown. The one cross section shown on the NDMCE base map cannot be checked against the SLRI plot because our cross section line is not draxvn that far into the project site. Was it ever that long? Yes, per HEC-2 input, NDMCE shows the cross section correctly. · FEMA requests that the stationing along the cross section be shown on the work map so that the GR data can be easily checked. NDMCE cross section line does not show any stationing. · Maps are slightly confusing because the proposed project and the cross section lines are not shown on the same map. Since maps are at a different scale, it is hard to determine whether the one cross section crosses the proposed improvements or just borders them. The proposed project does not show against the existing and proposed delineations either, therefore the proposed delineations do not make any sense. · Only cross section plotted is 93810. Is it the only effective model section on site? Yes. Was any cross section added? no. · Maps do not use most recent current aerial topographic maps. Maps based on NDMCE field surveys. Appropriate if contour interval is more accurate (yes: NDMCE: 1 ft, SLRI used TRFS maps: 2 ft) or survey more accurate. Does it show features missed by aerial topographic maps? Not really. How does it compare with aerial topographic map? Favorably, it shows more detail. 5) Benchmarks · Maps do not refer to an existing FIS benchmark: There is no FIS benchmark nearby, not even on the adjoining Carrollton FIRM map. · Maps refer to nearby USC&GS benchmark ~ Carrollton dam. Same BM as SLRI? No, we used TRFS-18 and TRFS 17. Did we tie it? Yes, recently, for the Andre~v Brovm Park West CLOMR. Benchmark 7Y-1924-R35 was tied at elevation 438.531 using GPS. NDMCE uses elev. 438.54, USC&GS quad map shows 439. Therefore, The maps appropriately and accurately refer to the USC&GS benchmark at the Carrollton dam over the Elm Fork of the Trinity River. · Small typo needs clarification: "Carrollton dam on East Fork Trinity river" should read "Carrollton dam on East side of the Elm Fork of the Trinity river". Inconsequential. 6) Duplicate effective model - Multi-profile. · NDMCE used Elm Fork models because site is subject to backwater effects rather than Denton Creek flows. · Models were obtained from FEMA, EFFISDPM.IH2 (multi-profile) and EFFISEXM.IH2 (floodway). They are different from the ones we used for SLRI in June 1994. The SLRI model included the comment that it (the SLRI model) xvas a shortened version of the Corps' reconnaissance model, modified per FEMA suggestions, ~vith discharges values higher than those in the NDMCE model and in the tis effective model. · Discharges are different from the SLRI duplicate effective model. At the first cross section in NDMCE model, Q's are 24300, 37100, 44350, and 60400 respectively. SLRI Q's were 24400, 39900, 48600 and 91200 (submitted in 1994). · N-values are identical to the SLRI ones at beginning of NDMCE model · Flow lines are similar except upstream of NDMCE project site, 98450 - 99280. · NDMCE duplicate effective model does not include cross section 98650. That would indicate a change by FEMA though, not NDMCE. · Cross section GR data different as well. · Why were there 2 different effective condition models in the first place? The effective model used in the SLRI should not even be the one used. Why isn't the SLRI proposed model used as the new effective condition model since the SLRI CLOMR was approved by FEMA on May 25, 19957 Is NDMCE using an older FEMA model that was the effective one at some time? This means that the Sandy Lake Road bridge that soon will be constructed across the Elm Fork of the Trinity River is mentioned in the text of the NDMCE report but not included in the model. Of course it also means that since all the models in the NDMCE report are equally affected by this condition, the NDMCE conclusion that this project only causes a 0.00 rise in water surface elevation is still valid (assuming all other hydraulic considerations in this report are correct). The proposed NDMCE model just cannot become the official effective model for FIS hydraulics. This issue should be cleared up by FEMA. Also, the project site would have been shown almost entirely in an ineffective flow area if the Sandy Lake Road and Mclnnish Park Improvements CLOMR had been used. · The 0.00 rise is based on the HEC-2 outputs for multi-profile models, NDMCE proposed model vs. FEMA effective model. If one looks at the table included in the report after Form 4, and if one chooses to compare proposed conditions with revised existing conditions, the project shows a .02 fl rise (the table is based on floodway models, not multi-profile models) - see above, 3rd item on page 2, and below-, 2nd item on page 7. · The SLRI CLOMR was based on an Elm Fork Hydraulic model obtained from Ron Morrison, P.E., the consulting engineer who prepared the City of Coppell FIS update. This is the model utilized to update the City of Coppell's FIRM dated April 15, 1994. · Other models are currently available for the Elm Fork of the Trinity River. The SLRI fully developed condition model was based on the best data that was available at the time (1994), the U.S. Army Corps of Engineers Trinity River Reconnaissance Study model, EF-CDC. Since then, the U.S. Army Corps of Engineers (USACE) finished developing a completely updated model that is intended to supersede EF-CDC and other older models of the Elm Fork of the Trinity River. The Upper Trinity River Feasibility Study models were received in our offices in April 1996, shortly after NDMCE performed sur~'eys for this project. They would be appropriate for the fully developed watershed analysis of the site. Future conditions discharges have also been computed by the USACE. 7) Revised Existing Conditions model - Multi-profile. · No cross section was added for this project but the one cross section that crosses the site was modified for existing conditions. · Thc body of thc text mentions that dimensions were taken from a topographic map (which one? Thc 1"=200' map included on thc flood plain work map looks like the TRFS topographic mapping) but one map indicates that thc topography was revised per field sur~'eys, which is O.K. The 446 contour on thc west side of thc property seems to coincide more with the 445 contour per NDMCE surveys, which might simply reflect thc greater accuracy of thc NDMCE map. · Thc revised GR data generally agrees with the existing contours on the work map. The ditch between thc 2 main land masses on thc site was made ineffective in the HEC-2 model (sta 9617- 9714 of cross section 93810), a decision that can be argued against, specially since that ditch will bc filled in thc proposed condition model and replaced by another swale 700 ft to the cast, that one fully (and somehow too) effective. Also, thc elevation 444.9 at station 9714 might have been intended to bc 443.9 to match the other side of thc ditch. · The revisions to the natural grade do not reflect the elevation of about 446, 200 ft upstream of the cross section, where the majority of the fill will be located. To better model the proposed fill, cross sections could have been added, through the middle of the site and where the fill ends near Denton Creek. As it is, the next cross section "seen" by HEC-2 is 3760 feet upstream, and small changes for this project have long-lasting effects in the computer model. One must consider, though, that the Denton Creek channel is ignored by HEC-2 in this area and that adding cross sections could be the equivalent of opening Pandora's box. Elevations around 444 instead of 446 would be more representative of the topography between cross sections 93810 and 97570. Such a model just does not reflect the proposed improvements with enough accuracy. · Results: Changes effected in the revised existing condition model have very little effect on the hydraulic model. Discharges through the right overbank decrease slightly (800 cfs), flow velocities, flow topwidth and water surface elevation are essentially unchanged for the base flood. The cross section is extended by 0.21 feet, similar to the 0.22 feet in the effective model. N- values were not modified. There is no dip in water surface elevation in the area. There are no abnormalities in the table 150 output data. Changes to the model did not create any special flow condition nor did they cause any new warning or error message to appear in the HEC-2 output. 8) Proposed condition model. · No cross section was added for this project but the one cross section that crosses the site xvas modified for proposed conditions. As mentioned above, since the cross section is not shown on the same map as the proposed improvements and both base maps are at different scale, it is hard to determine if the GR data was input correctly and whether the proposed delineations make any sense. · The proposed model "sees" very little of the proposed changes. Elevation of most of the proposed fill from 447 to 449, modeled as 448.5 (not much error, except that the next cross section is 3760 feet upstream). · On the proposed grading plan, the swale narrows down from about 250 ft wide to about 116 ft wide (bank width) as it flows north, its base varies from 146 ft to 190 ft to 70 ft and it is approximately 380 feet long. HEC-2 sees a 248 ft wide trapezoidal channel, from sta 8880 to sta 9128 (bank width), with a base 164 ft wide, from sta 8912 to 9076. The next cross section is still 3760 feet upstream. · Existing ponds on the North side of the site are re-graded. They were ineffective in the existing model and are still ineffective in the proposed model. · Because no cross sections were added, the mild grading at the extreme north end of the site is not modeled, just as the mild grading at the ponds is not modeled. · The swale does add conveyance in this model. It replaces a smaller ditch that is filled as part of this project but not modeled in existing conditions. The conveyance added might be temporaD' only, as the proposed Sandy Lake Road bridge that will soon be built will block off the 100- year probability flood on the right overbank of the Elm Fork. · But the cross section modified corresponds to cross section 93890 (cross sections were completely redefined) in the SLRI proposed model. The effective floxv at SLRI cross section 93890 extends on the right overbank as far west as station 9100, as is shown on the second attached Halff Associates, Inc. map. If the SLRI model had been used as the effective model, the property would have been shown almost completely in an ineffective flow area, where grade changes are irrelevant. · Results: Changes effected in the proposed condition model have very little effect on the hydraulic model, maybe because the bulk of the changes is not perceived by HEC-2. Discharges were not modified by NDMCE to model the proposed improvements. N-values were not modified. The proposed channel flow velocity is slightly lower than the existing one. Although the flow topwidth was reduced by 1288 ft (from 10032.21 ft to 8743.95 ft), water surface elevation increased by only 0.01 ft at cross section 93810 for the base flood. Flows on the right overbank increased by 2560 cfs. There is no dip in water surface elevation in the area. There are no abnormalities in the table 150 output data. Changes to the model did not create any special flow condition nor did they cause any new warning or error message to appear in the HEC-2 output. 9) Floodway models. · A floodway per FIRM map is shovm on the flood plain work map, it is similar to the one shown on the actual FIRM map. · Because the floodway and the proposed project are not shown on the same map, one cannot really tell in which way the floodway is impacted by the project. Is the top of bank line along Denton Creek the same on both maps? If so, the project cuts into the Denton Creek channel at a location on the north west side of the site, maybe at the outfall area, and almost certainly at 3 places next to the ponds. · The proposed improvements cause no change in water surface elevation. A possible reason why the proposed project has little effect on the floodway model is that no cross section in the floodway model reflects the true extend of the proposed changes. The model, without enough cross sections, is simply blind. · The data shown in the water surface elevation check table is identical to the one in the HEC-2 output for all floodway models. 10)Diffital models. · The digital models were received separately on September 3, 1996. No viruses were detected. · An on-screen comparison of files confirmed that only 1 cross section had been changed in all models, which agrees with the report. · The HEC-2 models were all run. The water surface elevations generated by the floodxvay models agreed with the data published in the report, for both the natural and the floodway runs. A small, irrelevant, discrepancy was found: the natural 100-year water surface elevations should be identical in the multi-profile and the corresponding floodway model. They are different by 0.01 ft for cross section 97570 (existing conditions) and for cross section 93810 (proposed conditions). Because the GR data is the same, the discrepancy can be attributed to the computer rounding numbers differently. 11)FEMA forms. · Form 4, 1.Reach to be revised: the upstream limit of the project does not correspond to a HEC-2 cross section. This is appropriate given the distance between the project and the next cross section upstream. · Form 4, 3.hydraulic analysis: this paragraph could mention the proposed improvements. · Form 4, 3.hydraulic analysis, models submitted: the existing model is more a corrected effective model than a pre-project conditions model because it incorporates new, more detailed topographic information. · Form 4, 4.model parameters: paragraph 4 should refer to the proposed condition model, not the existing one. · Form 4, 5.results: water surface elevation is higher than end points of cross section at 93810, according to the detailed output. The end of the cross section is at elevation 445 while the water surface elevation is at 445.22. · Form 4, 5.results: questions 2 through 5 should refer only to that portion of the model that was revised for this project. · Form 4, 6.flood profiles: a profile is not included with this report, although NDMCE should be right in assuming that since the water surface elevation is not visible, a profile revision is not necessary. · Form 4, 6.flood profiles: a floodway data table is actually included on page 20 of the HEC-2 output. · Form 4: The water surface elevation check table shows a proposed elevation of 445.23 at cross section 93810 while the HEC-2 output only shows 445.22. · Form 5, 1.mapping changes, 1.D: the stationing control is not indicated on the cross section revised. · Form 5, 1.mapping changes, 3: the revision request scale is shown at scale 1"=100' with 1 ft contour inter~,als, while the proposed grading plan is at scale 1"=60'. · Form 5, 1.mapping changes, 5.a: property owners have not been notified, which should be O.K. since the increase is insignificant and no structure is impacted. · Form 5, 2.earth fill placement, 3.C: fill has not been constructed yet, since this is a CLOMR, it has not been compacted yet and no structures could be placed on it. 12)Possible YEMA reply. · FEMA might decide that cross sections should have been added to the models. · FEMA will look for increases in water surface elevation greater than 0.00 ft within the floodway area, and for increases greater than 1.00 ft within the floodway fringe. If one accepts the model without any added cross section, those requirements are met. · FEMA will check that topwidths are consistent between the model and the work map. Since only part of the cross section is shown, they will have problems answering that question. · F£MA might argue against the "extended cross section" message in the HEC-2 output. The same message appears in the effective model though. · FEMA might have a hard time finding the cross section line on the flood plain work map. I know I did. They have generally not complained about not finding stations along cross section lines on maps. They should object to the absence of the cross section line on the proposed site plan. · FEMA could decide that the effective model is the Sandy Lake Road Improvements model and not the one used by NDMCE. · Irregularities in filling the forms are minor and should not cause any serious problem with FEMA.