Technical Report CERC-95-2 March 1995 US Army Corps of Engineers Waterways Experiment Station Physical Model Study of Revere Beach, Massachusetts by Donald L. Ward Approved For Public Release; Distribution Is Unlimited CERC-95-2 Prepared for U.S. Army Engineer Division, New England The contents of this report are not to be used for advertising, publication, or promotional purposes. Citation of trade names does not constitute an official endorsement or approval of the use of such commercial products. & PRINTED ON RECYCLED PAPER Technical Report CERC-95-2 March 1995 Physical Model Study of Revere Beach, Massachusetts by Donald L. Ward U.S. Army Corps of Engineers Waterways Experiment Station 3909 Halls Ferry Road Vicksburg, MS 39180-6199 iy il | | 7 Mh Final report Approved for public release; distribution is unlimited U.S. Army Engineer Division, New England Prepared for 424 Trapelo Road, Waltham, MA 02254-9149 US Army Corps of Engineers Waterways Experiment Station COASTAL ENGINEERING RESEARCH CENTER FOR INFORMATION CONTACT : PUBLIC AFFAIRS OFFICE U. S. ARMY ENGINEER WATERWAYS EXPERIMENT STATION 3909 HALLS FERRY ROAD VICKSBURG, MISSISSIPPI 39180-6199 PHONE : (601)634-2502 h = y }— STRUCTURES = LABORATORY AREA OF RESERVATION = 2.7 sqkm Waterways Experiment Station Cataloging-in-Publication Data Ward, Donald L. Physical model study of Revere Beach, Massachusetts / by Donald L. Ward ; prepared for U.S. Army Engineer Division, New England. 69 p. : ill. ; 28 cm. — (Technical report ; CERC-95-2) Includes bibliographic references. 1. Sea-walls — Massachusetts — Revere Beach. 2. Engineering mod- els. 3. Coastal engineering — Massachusetts. |. United States. Army. Corps of Engineers. New England Division. II. U.S. Army Engineer Wa- terways Experiment Station. Ill. Coastal Engineering Research Center (U.S.) IV. Title. V. Series: Technical report (U.S. Army Engineer Water- ways Experiment Station) ; CERC-95-2. TA7 W34 no.CERC-95-2 Contents Pe LACE? cen ee Ee Oe ee Se ee Rated Sech rari we v1 Conversion Factors, Non-SI to SI Units of Measurement ........ Vii 1 Introduction es ee ae ee ee eae 1 heyPrototy Pers ets woes co) cule et Serotec ent as eee 1 MNEsPLODS II rte errs ee ee ee Ee ee, Rn pe ek S 2) SCOPCHOLAWOlkK reat cack mien me ae enr ar er ee ire tS setae ce neneen anes 4 D=MeStAacilityenwe tae eae Chern eee Ce ae ee Cet a S) 3=—Researchwbasks An BavandyGyaren toe tend eee sera ee ey ae 9 aA SKGAW Reuter eed RR ae ete ete aT On aE 9 SIRENS SC baie eet sie cue Aue oe a ee aA etal Mae a tales MEN aaa anor an clad fo 16 MAS Ke eee av ac F eta Pulee ae sa OE ee Rone eR Dal ResressionyAnalysis fsa. hoes eho ns ee oe ee 28 4—Discussion of Research Tasks A, B,andC .............. 44 5—-Revere Dike Study 27 32 cee fa kee Od eee Re 46 ParkoDike: OOM Profile aces cottesict cn oe tee nen or ae ee 46 Rubble-Mound Dike, 1991 Profile................... 50 Parks Dike 97S'Protiles tne See SU Se eRe Sys 52 Rubble-Mound Dike, 1978 Profile................... 54 ArmornUnitiStabilityonr econ cast oth ny cigs a rece sto eR vio: 56 6—Revere Dike Discussioni eo ey ee eee eee ee 58 INC FETETICES! See wee Scien teeter ley Rape in be ranma ite fest tlre 61 SF 298 List of Figures Figure l., ‘Locationsmap)< 22... eat tel aetio tere Sir Meue es on le 2 Figure 2. Figure 3. Figure 4. Figure 5. Figure 6. Figure 7. Figure 8. Figure 9. Figure 10. Figure 11. Figure 12. Figure 13. Figure 14. Figure 15. Location of profiles used in SBEACH simulations. ... . 3 Plan and profile views of 18-in. wave flume ........ 6 Plan and profile views of 3-ft wave flume.......... 7 Storm profile for storm of February 1978 .......... 10 Profile 2 as surveyed in 1978 and as reproduced forphysicalsmodelistudyjas sisal ie acne 14 Storm profile for SPN from storm of November, 1945 .. 17 Profile 1 as surveyed in 1991 and as reproduced iniphysicalimodelistudy; 4) areca eee 20 Profile 3 as surveyed in 1991 and as reproduced injphysicalemodelistudyiieaeienceene ene nna anne 21 Profile 4 as surveyed in 1991 and as reproduced injphysicalemodeltstudy Myers ere eNL ei ean a: een aeie D2 Profile 5 as surveyed in 1991 and as reproduced injphysicalimodeléstudyaar ene eee 23 Profile 2 as surveyed in 1991 and as reproduced imuphysicalimodelistudy Mea ieee ee oc eecne 25 Portion of prototype plans for the proposed park dike ... 47 Crossisectiontofimodell parkidikeyy sii een 48 Cross section of rubble-mound dike ............. 50 List of Tables Table 1. Table 2. Table 3. Table 4. Table 5. Wave Data from Numerical Model SBEACH and Inter- polated Wave Conditions Tested for 1978 Storm at Revere Beach: Profiler 2c oi ape eee ect pete ee ue 11 Revised Wave Data from Numerical Model SBEACH and Interpolated Wave Conditions Tested for 1978 Storm atpRevereyBeachserohiley2a een eae er ae 12 Overtopping for 1978 Storm at Profile 2 for Prototype Seawalliensthiois3:890 ita wee eae 15 Wave Data from SBEACH and Interpolated Wave Conditions with 1991 Profiles at Revere Beach ...... 18 Representative Seawall Crest Elevations for Overtopping Studysof 1991)Beach Profiles; .. aya renee eae 24 Table 6 Table 7. Table 8. Table 9. Table 10. Table 11. Table 12. Table 13. Table 14. Table 15. Table 16. Table 17. Table 18. Table 19. Overtopping Rates and Volumes for SPN with 1991 BeachiProfiles erases can Gs ces i ee ee ee ee 26 Test Conditions and Overtopping Rates for TaskC .... 29 Wave Conditions and Seawall Elevations for 1991 Profiles OfREeVveresB cachet aes ei caries Aen iar Rec eer 33 Minimum and Maximum Values for Parameters Used in Regression Analysis of 1991 Overtopping Rates .... 35 Input Data for Revere Beach Overtopping Rates Regression Analysis, Task C, Profile No. 2, Survey Year 1978 .... 37 Minimum and Maximum Values for Parameters in the Regression Analysis for TaskC ............... 38 Input Data for Revere Beach Overtopping Rates Regression Analysis, Profile No. 2, Survey Year1978 ......... 40 Minimum and Maximum Values for Parameters in the Regression Analysis with 1978 Profile ........... 42 Minimum and Maximum Values for Parameters Used in Regression Analysis of Combined Data from Tasks A, Beer Geet ir, gree OA, MecATS ht Neon tie Vey Tete Aon 43 Test Conditions and Overtopping Rates for Park Dike with 199 Profuless 2/270 oh Sea Aee, by Ween End vik APE 49 Test Conditions and Overtopping Rates for Rubble-mound Dikerwithwl99 1 Profile eegaeg = ae eet rary 51 Test Conditions and Overtopping Rates for Park Dike WithplO7SyProtile; mies 6. crus ches Reo ccae atm ese 53 Test Conditions, Overtopping Rates, and Armor Stone Displacement for Rubble-mound Dike WithplO7 Si Profile eos cect Reich cas date) fsa ss ee air oe ws 55 Comparison of Overtopping Rates at Peak of SPN .... 60 Preface The physical model study of Revere Beach, Massachusetts, reported herein was requested by U.S. Army Engineer Division, New England (CENED), as part of the Saugus River and Tributaries Flood Damage Re- duction Project. The investigation was conducted at the Coastal Engineer- ing Research Center (CERC) of the U.S. Army Engineer Waterways Experiment Station (WES) between December 1992 and July 1993. The physical model study was intended to provide overtopping data and empiri- cal equations that could be incorporated into numerical model coastal pro- cesses, to aid in the design of a new dike to provide flood protection for the city of Revere, and to supplement a coastal processes study conducted by personnel in the Coastal Oceanography Branch and Coastal Processes Branch of the Research Division at CERC (Smith et al. 1994). This study was conducted by personnel of CERC under the general direction of Dr. James R. Houston, Director, CERC, and Mr. Charles C. Calhoun, Jr., Assistant Director, CERC. Direct supervision was provided by Messrs. C. E. Chatham, Chief, Wave Dynamics Division (WDD), and D. D. Davidson, Chief, Wave Research Branch (WRB), WDD. This report was prepared by Messrs. Donald L. Ward, Principal Investigator, WEB, and John M. Heggins, Computer Technician, WRB. At the time of publication of this report, Director of WES was Dr. Robert W. Whalin. Commander was COL Bruce K. Howard, EN. The contents of this report are not to be used for advertising, publication, or promotional purposes. Citation of trade names does not constitute an official endorsement or approval of the use of such commercial products. Conversion Factors, Non-SI to Sl Units of Measurement Non-SI units of measurement used in this report can be converted to SI units as follows: Multiply | By To Obtain acre-feet 1,233.489 cubic feet per second 0.02831685 0.7645549 cubic yards cubic meters cubic meters per second cubic meters meters centimeters a! 2.54 1.609347 nautical miles 1.852 pounds (mass) 0.4535924 16.01846 pounds (mass) per cubic foot kilometers kilometers kilograms kilograms per cubic meter Vii ; congo ie aA bait yi serge aig eppanl4 Waa, 1 Introduction The Prototype Revere Beach in the City of Revere, Massachusetts, is located on the Massachusetts coastline approximately 6 miles! northeast of Boston (Fig- ure 1). The beach is located on Broad Sound, bordered by Roughans Point headland to the south and Point of Pines to the north, and is partially sheltered by the Nahant Peninsula to the northeast. Revere Beach is the oldest public beach in the nation, with boundaries established in 1895. Erosion of the beach led to construction of protective seawalls in the 1920’s along most of the reach. The seawalls were not sufficient to pre- vent severe flooding of backshore areas. The beach continued to erode, and waterfront establishments suffered from the floodings. Approximately 600,000 cu yd of fill were placed along Revere Beach in 1991 as part of the Revere Beach Restoration project constructed by the U.S. Army Engineer Division, New England (NED). The fill placed a 50-ft-wide berm in front of the seawall at elevation 18.0 ft mean low water (mlw), which is typically 2 to 3 ft below the seawall crest. Although the beach was not designed or justified for flood level reduction, it does reduce wave overtopping, resulting in incidental flood reduction benefits. Eight bathymetric profiles were surveyed along Revere Beach and Point of Pines after a 1978 storm, extending offshore from the seawall for approximately 10,000 ft (Figure 2). The surveys were repeated after a beachfill project in 1991, and a storm which followed in October 1991. 1 4 “5 Pee A table of factors for converting non-SI units of measurement to SI units is presented on page vii. Chapter 1 Introduction Revere Beach Revere REACH Nahant Boston Roughans Point Figure 1. Location map The Problem As part of the Saugus River and Tributaries Flood Damage Reduction Project, a flood control plan includes construction of a tidal floodgate across the mouth of the Saugus River, and new walls, dikes, revetments, and dunes along the shorefronts of Revere and Lynn. Also, land is being purchased along the Saugus and Pines Rivers to provide a holding area for any floodwater that may overtop the beach and seawall as well as retain- ing rainfall runoff during floodgate closure. The problem was to deter- mine the amount of overtopping expected, to assist NED in determining the quantity of land to be purchased. Wave hindcast studies were con- ducted by the Research Division of the U.S. Army Engineer Waterways Experiment Station’s Coastal Engineering Research Center (CERC) to de- termine major storm conditions at offshore locations in the area, and nu- merical models were used to propagate storm waves shoreward to the beach. The numerical models could not adequately predict overtopping along the beach, and CERC’s Wave Dynamics Division was requested to determine the overtopping rates through physical model studies, using wave and water level data supplied by the numerical models. Information on the physical model studies is contained in this report. For information on the numerical model studies, see Smith et al. (1994). Chapter 1 Introduction Residential Area Saugus | River Revere never® LEGEND [) Beach at High Tide Beach St- [2] Wetland Roughans — Seawall O Baie ess Revetment Elliot om Dunes Circle 8 Profile Number {e) 1500FT Ee! Figure 2. Location of profiles used in SBEACH simulations To further protect against flood damage, a dike has been proposed for construction on the west side of Revere Boulevard along a reach of Re- vere Beach, and the state is interested in developing the dike into public parkland. Overtopping values are needed to design the dike, but could not be accurately estimated by analytical techniques due to complexities of wave action flowing over the beach bathymetry, overtopping the seawall, crossing Revere Boulevard, overtopping a toe wall at the seaward edge of the dike, and then flowing up the “park” dike. CERC was therefore asked to determine overtopping rates for the proposed park dike through physical model tests. To assist NED in determining the level of non-federal fund- ing, CERC also was asked to determine a minimum dike configuration, Chapter 1 Introduction rubble-mound structure for flood control only, that would provide protec- tion against overtopping along Revere Boulevard. Scope of Work The physical modeling study was divided into four tasks. Task A was to confirm the validity of numerical and physical models by recreating a known storm event and comparing the overtopping in the physical model to measured overtopping in the backshore, using the beach profile that ex- isted at that time. Task B was to determine overtopping for the duration of the selected design storm event along Revere Beach using beach pro- files surveyed after the beachfill project. Data from Task B also were used by CERC’s Research Division to develop a “bore runup overtopping module” for Revere Beach to be used with numerical models. Task C was to use one of the profiles taken before the beachfill project and determine overtopping rates for storm conditions selected from a synthetic storm database developed from major storm events identified by wave hindcast- ing. Data from Task C were used by CERC’s Research Division to de- velop a “broken wave overtopping module” for Revere Beach to be used with numerical models. The fourth task of the physical modeling effort was the study of the pro- posed dike along Revere Boulevard. The study included determination of overtopping rates for the proposed park dike and a rubble-mound dike for flood control only when fronted by profiles from both the 1978 and 1991 surveys and subjected to severe storm conditions. Chapter 1 Introduction 2 Test Facility Two-dimensional (2-D) physical model tests were conducted in CERC’s 150-ft-long by 1.5-ft-wide by 3.0-ft-deep wave tank (“18-in. flume”) and 150-ft-long by 3.0-ft-wide by 3.0-ft-deep wave tank (“3-ft flume”). In both flumes, waves were generated by a piston-type wave board powered by an electro-hydraulic pump controlled by a computer- generated signal. The 18-in. flume has an existing 1:30 (V:H) concrete slope starting 60 ft from the wave board; the 3-ft flume has a 1:20 con- crete slope starting 36 ft from the wave board and extending for 10 ft, fol- lowed by an approximately 1:100 slope. Pre-test conditons of the 18-in. and 3-ft flumes are illustrated in Figures 3 and 4, respectively. The flumes were modified for each test to meet specific profile needs. The models were built to a geometrically undistorted linear scale of 1:20 (model:prototype) for Task A and 1:30 for Tasks B, C, and the dike study. Based on Froude’s model law (Stevens et al. 1942), the following model-to-prototype relationships were derived. Dimensions are in terms of length L and time T. Model-to-Prototype Scale Relation Characteristic Dimension Area fiZ Ar = (L)* ae Volume V, = (LA)? 1:27,000 1:5.477 Water that overtopped the seawalls during physical model tests was pumped into a rectangular catch basin at the conclusion of the test run. The change in elevation of water in the catch basin was then measured with a point gauge and converted to prototype overtopping rate in cubic feet per second per linear foot of prototype seawall (cfs/ft) by the follow- ing relationship: Chapter 2 Test Facility OWN} BACM “UI-8} JO SMAIA OII}O1d pue Ue; MalA 3a11d0¥"d ‘1334 NI 3YV SLINAWSENSV3AW T1V ‘AS = Hb -A1VOS G3LYOLSIG 00st —_—__——“_- i _ ——————— OLS as sa YOLVHSANAD SAVM SdAL-NOLSId MalA NW1d ‘1334 NI Suv SLINSWSYNSVAWN TIV ‘AS = HE -AWWOS G3ALYOLSIC "€ aunBi4 Chapter 2 Test Facility QUIN} SAEM Y-E JO SMAIA BIIJO1d pue UR|q “py OiNBI4 MalA 411d08d 1334 NI Suv SLNSWAYNSVAW TIV HOL=AL ‘S1VOS GaLYOLsid 0st <—_______—__ #9. —______—___p»> 001 pae——__ 996 —— SI 02:4 OOL:+ YOLVHSNAD SAVM AdAL-NOLSId MalA NW 1d 1534 NI SYV SLNSWSENSVAW TV HOL=AL ‘SIVOS GSLYOLSIC OELL r — —_ | | PF- |i ok ~~ Chapter 2 Test Facility OT = prototype overtopping rate in cfs/ft A-p = cross-sectional area of model catch basin, ft? S = scale factor, either 20 for 1:20 or 30 for 1:30 H,.», = change in water surface elevation in the model catch basin, ft T,y4n = time of model test run, sec run Wrume = Width of model test flume, ft For each set of tests, the cross-sectional area of the model catch basin, flume width, time of model run, and scale factor were constants. The over- topping rate could therefore be calculated as 3/2 A.» x § P| TARA ony 2 n ume ru where the first term is a constant and the only variable is change in eleva- tion of the water surface in the catch basin. Chapter 2 Test Facility 3 Research Tasks A, B, and C Task A Purpose The purpose of Task A was to validate the numerical and physical models by reproducing the effects of a known storm event. The storm selected for the test occurred in February, 1978; surveys of high-water marks in ponding areas provided an estimate of the total overtopping over the reach represented by Profile 2, which is located fronting the proposed park dike. Using wave data supplied by CERC’s Research Division, Task A simulated the 1978 storm in the physical model to determine if overtopping measured in the 2-D model corresponded to the estimated prototype overtopping volume. Selection of test conditions The National Ocean Survey (NOS) Boston Harbor tide gauge provided data on local water levels. Wave conditions were determined by CERC’s Research Division with the following numerical models: deepwater wave hindcast information was brought shoreward using numerical model SHALWYV, diffracted and refracted into Broad Sound by numerical model REF/DIF, and transformed shoreward along each of the surveyed profiles using numerical model SBEACH. SHALWYV uses spectral wave informa- tion; REF/DIF and SBEACH use monochromatic waves. The height of the average of the one-third highest waves (significant wave height H,) and the period of peak energy density (7,,) calculated from the wave en- ergy spectrum determined by SHALWV were chosen to characterize the monochromatic wave transformation in REF/DIF and SBEACH. Still-water level (swl), wave height, and wave period were determined for each hour during the selected storm to define the storm profile, and test conditions for the physical model study were selected from the storm profile. Figure 5 shows the storm profile used as input to SBEACH. For the physical model study, water level at the peak of the second tide cycle Chapter 3 Research Tasks A, B, and C 10 x>) (é%) as @ = a (e) — Lye) N ql For the second set of tests, the swl at hr 30 was chosen for the lowest water level (11.0 ft mlw), and the swl at hr 14 was chosen for the higher water level (13.2 ft mlw). Similar to the first set of tests, linear interpola- tion was used to determine the time at which the swl to be tested occurred and the wave height and period at that time. Test conditions are shown in Table 2. Because of time restraints imposed by having to rerun the storm pro- file, the second set of tests was reduced to a single 20-min run at each of the four test conditions on the lowest water level (11.0 ft mlw), two 10-min runs at each of the test conditions at the next higher water level (13.2 ft Chapter 3 Research Tasks A, B, and C mlw), and five 2-min runs at each of the three highest water levels. As in the earlier set of tests, multiple runs of short duration were used at the highest water levels to allow the overtopped water to be returned to the flume to maintain the swl. Determination of model profile The existing 1:30 concrete slope in the 18-in flume did not match the beach survey taken after the 1978 storm. Therefore, an entirely new pro- file was constructed and installed seaward of the existing concrete slope. The beach profile was displayed on a computer screen and an idealized profile was determined by matching a series of straight lines to the actual profile as closely as feasible, including a horizontal line to use as the flume bottom. The actual profile and the idealized profile are shown in Figure 6. With the depth at the flume bottom determined, model scale was estab- lished by limitations of the wave generator. The wave generator was un- able to generate the required signals at scales larger than 1:20; therefore, the model was constructed at a 1:20 scale. The idealized profile was constructed of plywood and placed in the wave flume over the concrete slope. When the slope was within 0.75 in. of the flume bottom (thickness of the plywood), 20-gauge sheet metal was used to extend the slope to the bottom of the flume. A vertical seawall was placed at the top of the plywood slope. Water overtopping the sea- wall accumulated behind the seawall and was pumped into a separate can- ister for accurate measurement of the overtopping volume at the end of each test run. Results Overtopping rates for the first set of conditions tested are listed in Table 1; overtopping rates for the repeated storm profile are listed in Table 3. Data in the tables have been converted to prototype scale. To determine total overtopping during the storm, it was assumed that the overtopping rate determined for a given point in the storm profile was constant over the time period extending from half-way between the given point and the preceding point to half-way between the given point and the following point. Because data were available at every 1-hr interval of the storm, overtopping rates at the first and last points tested were assumed to exist for 1/2 hr before and after the point tested, respectively. Multiplying the overtopping rate for a tested point in the storm profile by the length of time the storm was assumed constant at those conditions yielded the vol- ume of overtopping for that test per foot of seawall, and multiplying by the length of seawall contributing to the flood zone yielded the total volume Chapter 3 Research Tasks A, B, and C 13 JOJCIBUSE) BAC/AA Apnjs japow jeoisAyd 40) peonpoides se pue 9761 Ul peAenins se Zz alljoid (4) a10YsHO eoueR}sIq (spuesnoy!) ol L wo}oq ewn|4 lepoW [ed!sAyq U! ajyO1q yoeag a[JO1q yoeag edAjo}o1g ———— BLEL ‘2 AOld adojs 0€:1 Bunsixy oo oOoOlUmeMmrmUmUmDELmUCCOWUmULULODLC HTC HON m oO < 9 = fe) =) ) (mjw "9 ainbi4 Chapter 3 Research Tasks A, B, and C 14 Table 3 Overtopping for 1978 Storm at Profile 2 for Prototype Seawall Length of 3,890 ft | Total Acre-ft for Begin Time Overtopping | 3,890-ft Hour Hour End Hour sec Rate, cfs/ft Length 13.29 12.50 13.64 4,094 0.0097 3.56 14 13.98 13.64 14.49 3,077 0.0994 27.32 15 15.00 140 [15.44 | 3,421 0.5200 158.88 16 15.88 15.44 16.73 4,642 0.1659 68.76 = 17 18 17.58 16. ar 18.50 6,366 0.0103 5.86 25.34 2aso |as7a | 74 4,458 0.0100 a 26.13 25.74 26.82 3,879 0.1674 57.99 27.50 26.82 27.50 2,463 0.8141 179.08 27.50 27.50 28.30 2,888 0.8141 209.93 rae 29.10 28.30 aos 4,500 0.1359 54.60 30.00 29.55 30.50 3,412 0.0116 3.54 Storm Total | 773 acre-ft of overtopping over the seawall for the time period that was tested. For this series of tests, NED determined that 3,890 ft of seawall would contrib- ute to the flood zone. Overtopping rates and volume for each hour of the storm are shown in Table 3. Based on surveys of high-water marks, NED calculated that about 600 acre-ft of water overtopped the seawall during the 1978 storm. The physical model test showed a total overtopping of 773 acre-ft, roughly 29 percent higher than the surveys had indicated. Due to uncertainties in the surveyed results, numerical models, and physical model tests, test re- sults were surprisingly close to the predicted results. Uncertainties in the tests are discussed in Chapter 4, “Discussion of Research Tasks A, B, and Cy Chapter 3 Research Tasks A, B, and C 15 16 Task B Purposes Purposes of Task B were to determine total overtopping for the design storm event for the beach profiles surveyed in 1991 (after the beachfill project) and to generate input data for a bore runup overtopping module to be used with numerical models by CERC’s Research Division. Using wave data supplied by CERC’s Research Division, Task B reproduced in physical models the five beach profiles located along Revere Beach and subjected them to the design storm event. Overtopping was measured for each profile at each hour of the storm tested. Selection of test conditions The design storm event, or Standard Project Northeaster (SPN), was based on a storm that occurred in November 1945. Wave conditions dur- ing the storm were obtained by hindcasting; still-water levels during the storm were obtained from the NOS Boston Harbor tide gauge. NED de- fined the SPN as the wave conditions from the storm profile determined by hindcasting for the November 1945 storm, but with an additional foot added to the swl recorded by the NOS tide gauge throughout the storm. The SPN was input by CERC’s Research Division into the numerical models listed under Task A to obtain storm conditions at Revere Beach. Figure 7 shows the storm profile for the SPN used as input to SBEACH. Condi- tions to be tested in the physical model were selected from the storm pro- file to include the worst conditions that occurred during the storm (hr 30) plus conditions at two lower water levels during both tide cycles shown in the storm profile (hr 27, 33, 40, and 45 for the lowest water level and hr 28, 32, and 43 for the higher water level). However, the static beachfill pro- file reduced overtopping to such an extent that no overtopping occurred during tests at the lowest water level; therefore, additional points from the peaks of the tide cycles were selected for testing. As in Task A, linear in- terpolation was used where feasible to adjust wave heights and periods to maintain a constant swl for tests of the incoming and outgoing tides in both tide cycles. Test conditions and the approximate hour of the storm represented are listed in Table 4 after shoaling in SBEACH to the approxi- mate location of the wave generator. As in the second set of tests in Task A wave setup was allowed to occur naturally in the wave flume, and the wave setup adjustment from SBEACH was not used. Determination of model profile Beach Profiles 1, 3, 4, and 5 were reproduced in the 18-in. flume at a geometrically undistorted scale of 1:30. Examination of beach surveys taken in 1991 indicated portions of the profiles could be represented by the existing 1:30 concrete slope in the wave flume and the flat bottom of Chapter 3 Research Tasks A, B, and C x—<—< Have Height (ft) +—+—+ Hater Elevation (ft, MLW) Have Period (s) Figure 7. Storm profile for SPN from storm of November, 1945 the flume. Shoreward of the 1:30 slope, sheet metal was used to repro- duce the steeper portion of the beach profile. A vertical seawall was placed at the top of the slope, and water overtopping the seawall was col- lected and measured to determine overtopping rates. Surveyed profiles and model representations of Profiles 1, 3, 4, and 5 are shown in Figures 8 through 11, respectively. Seawall elevations varied over the reach represented by each profile. A representative seawall height was selected for each profile except Pro- file 1; two representative seawall heights were selected for Profile 1. Selected seawall elevations are listed in Table 5. Beach surveys started at the foot of the seawall, and the elevation at the foot of the seawall was reproduced in all model profiles except Pro- file 1. The beach surveyed at Profile 1 measured an elevation of +21.0 ft mlw at the base of the seawall with a seawall crest elevation reported at +21.4 ft mlw, providing a freeboard of 0.4 ft. However, selected represen- tative seawall elevations for that segment of Revere Beach were +19.8 ft and +20.7 ft, both of which are lower than the beach survey. Because the reaches represented by both seawall elevations were significant, it was de- cided to conduct the Profile | test series twice, with one complete set at a seawall elevation of +19.8 ft and one complete set at a seawall elevation of +20.7 ft. For the first set of tests on Profile 1, the profile was modeled such that the beach slope extended to an elevation of +19.4 ft and then re- mained at a constant elevation until reaching the seawall, resulting in a freeboard of 0.4 ft. For the second set of tests, the same slope was used to Chapter 3 Research Tasks A, B, and C 17 Table 4 Wave Data from SBEACH and Interpolated Wave Conditions with 1991 Profiles at Revere Beach SWL Wave Wave SWL Tested |Interp Interp Wave |Interp Wave Hour ft, mlw Height, ft Period, sec |ft, mlw Hour Height, ft Period, sec Profile 5 a F 8.80 15.90 8.30 15.90 8.92 15.90 6.89 15.90 4.69 15.90 oa ae) @orN ian to for) ay wo ° cA a ite) ro) ive) o (Continued) 18 Chapter 3 Research Tasks A, B, and C a Table 4 (Concluded) Wave Wave SWL Tested |Interp Height, ft ES sec |ft, mlw Hour Profile 3 (Concluded) Interp Wave |Interp Wave Profile 2 27 10.0 9.7 15.9 10.00 27.00 9.66 15.90 28 13.4 11.0 15.9 13.20 27.94 10.89 15.90 31 15.0 11.7 15.9 15.00 31.00 15.90 13.2 11.0 15.9 13.20 32.00 — 15.90 43 13.5 7.4 15.9 13.20 43.18 15.90 44 11.8 6.6 15.9 45 9.6 AL 15.9 10.00 44.82 15.90 Profile 1 15.9 19 Chapter 3 Research Tasks A, B, and C Apnis jeapow jeoisAud ul poonpoidas se pue 1661 ul! peAsains se | allJOlg (j) al0USHO eouR)sIG (spuesnoy]) S‘0 (U6L'2-) woyog ewn|4 edojs oe: Bunsixy lapoy jes!sAyd U! a[jo1d Yoeeg alyO1g Yoeag adAjoj}olg L661 ‘} al}Old (mjwW Y) UONeAe|z Chapter 3 Research Tasks A, B, and C 20 JOPCEIBUBL) BAC/A\ Apnis japow jeoisAyd ul paonpoideas se pue 1.66} ul paAanins se ¢€ ajI}O1g (y) asousyo eoueisiq (spuesnou]) oh (y Zp'1-) woyog awn|4 lapoyy [eo!sAyd U! ajyosg yoRag allJO1q yoeag adAjojolg L661 ‘€ A1YOld adojs 0€: 1 Bunsixy m ® < = = 5 Se 3 E38 "6 aunbi4 - N and C B Chapter 3 Research Tasks A Apnjs japow jesisAyd ul pednpoides se pue 166} ul paAesins se vy ajljoig “O} aunbi4 (4) e10ysyo eouejsiq (spuesnou,) cl (¥.69'0-) WwoyYog awn|4 edojs 0€: 1 Bunsixg JO}JEIBUSE BALAK Pe uoneag|y (mw ¥) lapow feoIsAyd U! ajyo1d YOeag alyOldg yoeag adAjojold L661 ‘p alljOld Chapter 3 Research Tasks A, B, and C 22 (YU ZL L+ ) Apnjs japow jeoisAyd u! peonpoides se pue 166 ul peAsains se Gg 9aIIJ01g (4) a10ysyo eoue}siq (spuesnou,) Z0 woyoq awn) edojs 0€:} Buysixy m ® < oy = 5 = 3 = lapoW JeaIsAyd U! B|yO1d yoeeg alyOldg yoeag edAjo}O14 L661 ‘S 8[Old “LL eunBi4 o N Chapter 3 Research Tasks A, B, and C ie an elevation of +19.4 ft, then an exten- Table 5 sion was added to continue the slope to Representative Seawall ; : He H an elevation of +20.3 ft, again providing Crest Elevations for a freeboard of 0.4 ft. Overtopping Study of VERN ECan FOES The wave generator in the 18-in. Seawall flume was unable to reproduce wave con- Elevation ditions at Profile 2 at a 1:30 scale. fools Rather than change to a smaller scale, Profile 2 was reproduced at a 1:30 scale in the 3-ft flume. Similar to the 18-in. flume, the existing 1:20 slope in the 3-ft flume was matched to a portion of the surveyed profile, and the steeper profile shoreward of the 1:20 slope was constructed of sheet metal. Surveyed and idealized profiles for Profile 2 are illustrated in Figure 12. Profile No. Results Overtopping rate per linear foot of prototype seawall for each profile and each hour of the storm that had measurable overtopping are shown in Table 6. Physical model tests were not conducted on Profile 5 at hr 31, or Profiles 3 and 4 at hr 42. Volumes listed in Table 6 for these tests were ob- tained by multiple regression analysis using the other results listed in Table 6. Regression analysis is discussed below. Storm conditions for the SPN were considerably worse than during the 1978 storm, with greater water depths and wave heights and longer wave periods. Overtopping rates, however, were considerably less, attesting to the incidental effectiveness of the 1991 beach fill. Overtopping rates mea- sured in the wave flume for Profile 3 were surprisingly low, but incident wave heights for Profile 3 were lower than for the other profiles. NED confirmed that in the prototype, overtopping rates at Profile 3 appeared lower than at the other profiles during the October 30, 1991, storm, and the general trends observed in the wave flume agreed with observations of the prototype. The model did not test erosion of the 1991 beachfill during the SPN storm. Therefore, higher overtopping rates could be experienced as the beach erodes during the storm. Tests in Task C+ below show re- sults if the beach should erode to 1978 contours. Chapter 3 Research Tasks A, B, and C Apnjs japow jeoisAyd ul padnpoides se pue 166} ul paAanins se z ayijo1g “z} eunbi4 al (4) es0YysYO eoue)sIq (spuesnoy|) ol (us6'Z-) woyoq ewn|4 Joyesauayy BACAA adojs Bunsixg m 14») < se¥) ee (eo) = = 3 = Jepow jedisAyd U! B[jO1d yOeeg alyO1g yoeag adAjojolg L661 ‘2 alljOld Chapter 3 Research Tasks A, B, and C 26 Table 6 Overtopping Rates and Volumes for SPN with 1991 Beach Profiles Total Seconds | Overtopping Overtopping sec Rate, cfs/ft Volume, cf/ft Profile 5 29.00 28.50 29.50 3600 0.1947 30.00 | 29.50 30.50 3600 0.3928 Ue ena Ue 00 [oo | 31.44 3384 0.1729 Profile 4 29 28 28.00 27.50 28.50 3600 0.0073 43 43.06 |ee9 43.50 [ee | [ee | 0078 28 Profile 3 43.06 42.53 [eee | 50 3492 Ei 8 Profile 2 | am faw few [ow fase 31 fst.00 | 30.50 31.50 32.00 31.50 32.46 3446 42.00 41.50 42.59 3918 ' Determined by regression analysis. (Continued) Chapter 3 Research Tasks A, B, and C Table 6 (Concluded) Total Seconds | Overtopping Overtopping sec Rate, cfs/ft Volume, cf/ft 29.50 30.50 3600 =f 31.89 31.44 32.40 3446 42.00 41.50 42.53 3706 43 43.06 lees 43.50 3494 0.0364 127 Profile 1b 28 28.00 27.50 eae 3600 0.0215 Ee oe [o_[e [we foose [mw elem foe loo fom loom Task C Purpose The purpose of Task C was to reproduce a selected set of conditions from a database of synthetic storm events (see Smith et al., in prepara- tion). Data from Task C were used to develop a broken-wave overtopping module for use with numerical models of Revere Beach. Selection of test conditions CERC’s Research Division selected storm conditions that were ex- pected to produce overtopping from broken-wave runup. All tests were conducted on the model of the 1978 survey of Profile 2 in the 18-in. flume. Chapter 3 Research Tasks A, B, and C ai 28 Test conditions selected by the Research Division for testing are listed in Table 7 as Tests 1 through 30. The selected tests were separated into groups with similar water depths to allow multiple tests to be conducted without changing water level in the wave flume. Table 7 also lists the ac- tual test conditions used. The wave generator in the 18-in. flume was un- able to produce the wave conditions for Tests 1 and 6; therefore, these tests were eliminated from the test series. Tests 25 and 26 were identical after adjusting the water level; therefore, Test 26 was deleted. The remain- ing tests were completed. It was desired to perform a multiple regression analysis on the results of the tests to obtain a relationship among overtopping rate, wave height, wave period, and still-water level. Eight additional tests therefore were conducted to provide a better range of test conditions on which to base the analysis. The additional test conditions are shown in Table 7 as Tests 31 through 38. At the conclusion of Task C, the Research Division asked that storm conditions selected from the SPN be tested with the 1978 profile. These additional tests were analyzed separately from Task C, and are therefore referred to in this report as Task C+. Six conditions representing peak hours of the storm were selected for testing. The wave generator in the 18-in. flume was unable to produce the wave heights at these conditions; therefore, tests were conducted at the highest obtainable H,,,, for the given swl and T,,. Conditions tested and results are given in Table 7 as Tests 39 through 44. Results Results of the test series are given in Table 7. Regression Analysis Purpose Regression analysis was performed on results of the physical model tests to determine relationships among overtopping rates, swl, and wave conditions. Regression analysis was conducted on results of Task B for the bore runup overtopping module, Task C (without C+, Tests 39 through 44) for the broken wave overtopping module, combined results of tasks A, C, and C+ for a “worst case” analysis, and on the entire set of tests. The Statistical Analysis System (SAS), version 6.04, was used for the analysis. Chapter 3 Research Tasks A, B, and C Table 7 Test Conditions and Overtopping Rates for Task C Test No. —_ — | — wo me js+ | oOo N wo }lrm ]— a ie) wo wo = — — Oo Oo [) ye) NI (oe) ie) 1s) — =< © Oo = fo) oO foe) ie) = — o — = o>) N w — Lye) (>) Oo Oo @ = ie) iN) —_ =x or) Ss ss fop) a || 6 w _ i) ° no —_ Oo a =n ° N for) N oO 2° ié>) x pS (Continued) wo nh = oO N N BS = [e%) oO oO f=) a (8) - Chapter 3 Research Tasks A, B, and C 29 30 Rate, cfs/ft 0.4309 Method Dimensionless parameters were selected that were suitable for the numerical models for which the regression models were destined. Over- topping rate was presented as OQ [=] cfs/ft = L?T7! where 1) I overtopping rate ool Il — Il appropriate dimensional units Dimensional parameters affecting overtopping rate include the following: fl-)ft=L Bilt se ails Jet (a) iS 12 Chapter 3 Research Tasks A, B, and C T [=] sec = T g [=] ft/sec? = LT? where f = structure freeboard defined as height of the seawall crest above the swl b = beach freeboard defined as height of the beach at the base of the seawall above the swl d = water depth at the flat bottom of the wave flume H = wave height defined as the monochromatic wave height at a distance of 2,000 ft offshore (approximate location of wave generator in model flume tests and the wave height on which the physical model tests were based) T = wave period associated with the monochromatic wave height H g = gravitational acceleration Dimensionless parameters that may also affect overtopping rates include: coté d/d2000 where cot®@ = cotangent of the beach slope defined as cotangent of the slope from the base of the seawall to the swl d2000 = depth at a distance of 2,000 ft offshore Because the model profiles did not extend to the wave generator, there was a difference in depth between the wave generator in the flume (ad- justed for scale) and the actual depth offshore of Revere Beach. The ratio d/d2000 is the ratio of the depth in the flume (adjusted for scale) to the depth where the wave heights were determined from the numerical model. Figures 8 through 12 show where the flume bottoms were fitted to the beach profiles and illustrate the differences between depths in the flumes and depths on the surveys at the location of the wave generator. Because input wave information (wave height and period) was obtained from SBEACH at the approximate location of the wave generator (approxi- mately 2,000 ft offshore from the seawall), it was thought that the differ- ence in depths, d/d2000, could play a role in defining the overtopping Chapter 3 Research Tasks A, B, and C 31 32 rates. Depths at 2,000 ft offshore varied somewhat throughout the storm due to sediment movement; therefore, the depth determined by SBEACH for each hour of the storm was used for analysis. The profile in the flume, of course, remained constant. Task B All wave flume tests conducted for the SPN used a wave period of 15.9 sec. Because this value was a constant for all tests, it was not used in the analysis. Gravitational acceleration was therefore the only term available by which to nondimensionalize overtopping rate in time. All other parameters required only a length scale, and either f or H were rea- sonable candidates for the repeating variable. After trying both variables, it was found that results were somewhat improved by using f. After many variations and combinations of terms were tried, the dimensionless vari- ables that provided the best fit to the data were arranged as follows: QO’ = Og*f?)'” PI = b/f PI2 = H/f PI3 = d/d2000 PI4 = cot PIS = d/f Data collected in the physical model tests were converted to prototype scale for the regression analysis. Input data are shown in Table 8. Note that the last three lines in Table 8 give the input data for the three points in Table 6 determined by regression analysis. Examination of the residuals from one of the regression models that was tried indicated that higher-order terms were required (a residual is the difference between Q’ predicted by the regression model and measured Q’). Second-order terms (squares of the P/ variables) and higher were therefore added to the analysis. Regression analysis was conducted on the dimensionless variable Q’. Any negative overtopping rates predicted were set to zero, and results were converted to predicted dimensional overtopping rates. Model selec- tion was then based on the sum of squares of differences between ob- served and predicted overtopping rates. SAS assumes a null hypothesis that the coefficient of a term in the model is zero, then computes the probability that the null hypothesis is true. Only terms in the model with a low probability of having zero Chapter 3 Research Tasks A, B, and C Table 8 Wave Conditions and Seawall Elevations for 1991 Profiles of Revere Beach’ eons Storm |SWL Depth in |Height Hour ft, mlw aaa ft : Seawall Elev at Wave Crest Base of |Over- Cotan | 2,000 ft Period |Elev. Seawall |topping |Beach /|Offshore sec sec ft, mlw cfs/ft [Slope _| ft, mlw 10.0 Hoes | 69 7.40 -3.57 eee oe ag sla fee 16.59 10.40 -3.57 4 16.6 17.29 7.10 -3.57 4 15.0 15.69 9.60 15.9 -3.57 4 13.4 14.09 7.91 15.9 -3.57 eee 487 | [acs hes 20.5 (18.2 0.0000 |15.5 — |-2.25 Is [aa ft. 15.9 |205 |18.2 0.0022 |15.5 — |-2.23 o}wo }|o 3 be hea 50 Eos bes le eo bus oe fou fico fires feos fo “5.02 2 es 2 16.15 7.29 16.0 -5.04 ze as.4 15.59 9.10 14.0 -5.42 16.6 14.0 -5.41 14.0 -5.41 159 19.8 19.4 0.0571 |14.0 -5.41 14.0 -5.45 , fe fos 1 eee eee ; 1 is foo few fam 15.9 19.8 19.4 0.0000 |14.0 -5.45 1 13.4 15.59 9.10 15.9 20.7 20.3 0.0215 |14.0 -5.42 1 29 15.9 18.09 9.30 15.9 20.7 20.3 0.3105 |14.0 -5.41 2f9:30), pala (Continued) = 1 All measurements are prototype scale. Chapter 3 Research Tasks A, B, and C 33 Table 8 (Concluded) ioe Seawall Elev at Wave Wave’ /|Crest Base of |Over- Cotan {| 2,000 ft Profile |Storm |SWL Depth in |Height Period |Elev. Seawall |topping |Beach | Offshore No. Hour ft, miw |Flume, ft | ft sec sec ft, miw cfs/ft Slope |ft, mlw 1 30 16.6 18.79 8.70 15.9 20.7 20.3 [0.5093 _| 14.0 1 |sez_|is4__|rs89_ | 901 [159 1 [x2 __|139[re09 | a90 [i590 207 feos |oosee [140 [5.45 42 fsa [aso [oso ise [aos fier | tes | s57 bees lee leo leo ee ea ee es | coefficients (typically 10 percent for this study) were retained in the selected models. The model that best fit the data in Task B was: Q’ = -0.0190100 + 0.113943*PI1 - 0.074790* PII? + 0.114503*PI32 - 0.072397*P134 - 0.007017*P/4 + 0.000199* P14? - 0.006809*PI5 + 0.001601*PI57 While this equation was somewhat tedious, it fit the data with a correla- tion coefficient of 0.991 (R2 = 0.983), and the sum of squares of differ- ences between the overtopping rates (dimensional) and measured overtopping was only 0.074. There were 38 data points in the analysis; therefore, the average difference between calculated and measured over- topping was + 0.044 cfs/ft. It should be emphasized that regression models presented in this report are site-specific and are only valid at Revere Beach and within the range of conditions tested. The range of variables used, both dimensional and nondimensional, is given in Table 9. It seemed unreasonable to delete wave height (P/2) from the model, es- pecially when a correlation analysis revealed that Q’ was more highly cor- related with dimensionless wave height than any other single variable. However, there was a very high correlation between dimensionless wave height and dimensionless water depth (P/2 and P/5, 76-percent correlation), 4 3 Chapter 3 Research Tasks A, B, and C | Table 9 | Minimum and Maximum Values for Parameters Used in | Regression Analysis of 1991 Overtopping Rates | | Parameter Min Max SWL, ft mlw 10.00 16.60 frame ifaw i frees se Fsematieecwae fae te fcemoumnsee www Overtopping rate, cfs/ft PB 0.7878 0.9586 14.0000 19.5000 4.6730 0.0101 =| which was expected for depth-limited breaking waves, and effects of wave height were therefore reflected in P/5. A much simpler model provided a reasonable fit to the data and used only P/1 and P/S5 (beach elevation and water depth). Initial analysis of the data revealed that overtopping rates for hr 30, Profile 1, at both sea- wall crest elevations were exerting a very high influence on the simplified regression model. Because these overtopping rates were extreme and will not be found in other storms for which the regression model will be used, these two values were excluded from the analysis. The model was: Q’ = -0.036533 + 0.099865*PI1 - 0.062324*P/* -0.003554*PI5 + 0.001114*PI5? This very simple model fit the data with a correlation coefficient of 0.969 (R2 = 0.939), sum of squares of the dimensional errors was 0.0796, and average difference between calculated and measured overtopping rates was + 0.047 cfs/ft. The exclusion of beach slope in this simplified model Chapter 3 Research Tasks A, B, and C a8 36 was probably due to the smail range of the variable (14.0 to 19.5) and the relatively short distance that the slope was used in the wave flume. This simplified model was used by CERC’s Research Division for the bore runup overtopping module. Task C Data from Task C (excluding C+, Tests 39 through 44) were analyzed to determine a regression model for a broken wave overtopping module. Input conditions for the regression analysis (in prototype scale) are given in Table 10. Overtopping rate was nondimensionalized in the same manner in Task B, that is, as Q’ = Ol(g*f >)'/2. Other variables that were determined to be significant in the regression analysis were: PI = swi/f P12 = Hif P= bop where L, is deepwater wavelength defined as = 2 L, = (gh(2n)*T The model that gave the best results was weighted by wave height and is given as Q’ = 0.004162 - 0.007285*P/1 + 0.003252* PII + 0.001559* P12? - 0.000025997*PI3 + 0.000000217* PI3~ As in Task B, this model was selected based on the sum of squares of re- siduals of the dimensional overtopping rates. Correlation coefficient for the nondimensional model was 0.9865 (R7 = 0.9732). Sum of squares of residuals for the dimensional overtopping was 0.0511 for 35 test runs, yielding an average error of + 0.038 cfs/ft. It should again be emphasized that the regression analysis should not be used beyond the limits of the data set or for any other sites. Table 11 lists the ranges of variables used in the analysis. 1978 Profile All tests conducted using the 1978 survey of Profile 2 were combined in a single data set for analysis. The data included Tasks A and C plus the Chapter 3 Research Tasks A, B, and C Table 10 Input Data for Revere Beach Overtopping Rates Regression Analysis, Task C, Profile No. 2, Survey Year 1978 Overtopping Rate 0.2397 7.80 0.2589 3.10 0.1380 bh N (>) —_ —_ —_ —_ —_ a Fe Ne eee RS oO | @ N wo }o!]w | wo oln >} o/a /a/5 Colo SWL, ft miw Wave Height, ft Wave Period, sec cfs/ft | 14.90 10.50 9.00 eile Glos me wal 14.10 8.10 0.4331 14.10 9.00 10.70 0.6230 13.40 12.00 13.40 0.2530 13.40 0.2970 easy D8 sed Loe jase 3.10 0.2679 0.2001 7.60 0.1658 2.00 7.50 0.1109 0.0889 | 2.00 0.1098 —_ —_ —_ —_ —_ —_ —_ — —_ —_ = ne wo i) oo fo?) [o>) a a a l °o fo) °o o fo} 1.60 3.10 0.0278 7.10 0.0691 = aaa le = rN > |e jo la fo) fo} (2) w = oO oO oO oO (=) 7.30 0.0812 (Continued) Note: Seawall Crest Elev. = 21.0 ft mlw; Elev. Base of Seawall = 9.2 ft mlw; Cotan Beach Slope = 10.7; Elev. at Flume Bottom = -3.00 ft mlw. Chapter 3 Research Tasks A, B, and C 37 38 Table 10 (Concluded) _-_———— SWL, ft miw Wave Height, ft Overtopping Rate Wave Period, sec cfs/ft 10.70 = BS ° 10.70 5.50 10.70 6.70 9.50 6.50 14.10 6.00 14.10 6.00 0 13.10 5.10 13.00 2° oO w a B 9.00 0.0274 8.60 0.0314 14.10 0.0052 12.70 0.4309 9.00 0.2795 12.00 ay (¢) ae (2) 12.00 7.80 = a an ° fop) ° =a ° N ° Table 11 Minimum and Maximum Values for Parameters in the Regression Analysis for Task C Parameter SWL, ft miw 9.5 Wave height, ft 3.1 Wave period, sec Seawail freeboard, ft 11.5 Beach freeboard, ft -5.7 -0.3 Overtopping rate, cfs/ft 0.0052 0.7475 PH 0.826 2.443 Pl 0.330 1.721 133.239 Chapter 3 Research Tasks A, B, and C additional tests conducted after Task C listed as C+. The data set for this effort is given in Table 12. Dimensionless variables that yielded the best results were similar to those used in the regression analysis of Task B, above. Dimensionless overtopping was defined in the same way, and the repeating variable was again seawall freeboard (distance between seawall crest elevation and swl). Dimensionless beach elevation used in Task B was replaced with the dimensionless difference between seawall crest elevation and beach elevation, and P/J4 was deleted because beach slope for the 1978 profile was constant. Wave period was a factor, and was characterized by deep- water wavelength. Depth in the flume and sw differed by a constant; therefore, they could not both be used and swl was selected for PJ5. The dimensionless variables are listed below. Q’ = Ol(g*f*)'" Pll = (f-b)/f J 2 = ely PI5 = swi/f PI6=L/f The dimensionless variable P/3 (d/d2000) used in Task B was not in- cluded because data for d2000 were not available for conditions in Task C. In conducting the analysis, one point was found to lie outside the gen- eral trend. In the set of six tests conducted as Task C+, the measured over- topping from the test with an swl of 15.0 ft was substantially greater than predicted. This data point yielded an unacceptable influence on the re- sults, and was therefore deleted from the analysis. For the remaining data, the selected regression model is given below. Q’ = -0.000338 + 0.002530*PI1* - 0.004788*PI52 + 0.001912*P123 - 0.000322* P/2® + 0.000000212*P/62 - 6.92016*10°!2*PI64 This model was selected by weighting the analysis by wave period, thereby increasing the significance of longer period waves. This model had a correlation coefficient of 0.992 (R2 = 0.984), and the correlation coefficient of dimensional overtopping (measured to predicted) was 0.970. Sum squares of the residuals of dimensional overtopping was 0.3106 for the 60 tests; therefore, the average error was 0.072 cfs/ft. Chapter 3 Research Tasks A, B, and C 39 40 Table 12 Input Data for Revere Beach Overtopping Rates Regression Analysis, Profile No. 2, Survey Year 1978 Overtopping I SWL, ft mlw Wave Height, ft |Wave Period, sec|Rate, cfs/ft 10.82 6.14 8.10 0.0066 ——} 12.96 6.99 8.60 0.0643 12.96 10.05 9.70 0.1004 10.82 Stco Nn lttce 0.0080 floras nnd 6.00 0.0077 6.76 0.0843 1.3553 0.0959 0.0063 11.00 8.25 0.0097 13.15 8.59 0.0994 ie; Cc C ie; ie; Cc Cc Cc Cc Cc 13.40 7.60 12.00 0.3002 J Cc 13.40 7.80 9.00 0.2530 Het (Continued) Note: Seawall Crest Elev. = 21.0 ft mlw; Elev. Base of Seawall = 9.2 ft mlw; Cotan Beach Slope = 10.7; Elev. at Flume Bottom = -3.00 ft mlw. Chapter 3 Research Tasks A, B, and C [Table 12 (Concluded) Overtopping Task |SWL, ftmlw | ft mlw Wave | Wave Height, ft | ft |Wave Period, sec| Rate, cfs/ft 13.40 7.90 12.30 0.2970 13.10 7.70 11.30 0.2397 Cc Cc Cc Cc Cc Cc Cc ra [am hes "MES PE a 0.2679 Chapter 3 Research Tasks A, B, and C 41 42 As with all regression models presented in this report, this model is only valid for the range of conditions tested and for the specific project site. The range of variables, both dimensional and nondimensional, used in this analysis is given in Table 13. Table 13 Minimum and Maximum Values for Parameters in the Regression Analysis with 1978 Profile Parameter Min SWL, ft mlw Wave height, ft 3. Wave period, sec Seawall freeboard, ft 4 Max Overtopping rate, cfs/ft Moroes en 1.3553 PH 1.026 Pl2 0.330 PIS 0.826 3.773 PI6 33.029 294.454 Qa 0.0000 0.0229 Combined regression analysis for Tasks A, B, C, and C+ All data collected in Tasks A, B, C, and C+ were combined into a sin- gle data set to develop a general regression model for the overtopping at Revere Beach. The combined data set includes all data listed in Tables 8 and 12, with the exception of the outlier mentioned above under 1978 pro- file. The same dimensionless parameters used in the analysis of the 1978 profile were used in the current analysis, but P/4 was added to include the beach slope. The variables are therefore defined as QO’ = OW g*f >)! PI = (f-b)/f PI2 = H/f PI4 = cot0 PI5 = swi/f Chapter 3 Research Tasks A, B, and C PI6 = L Jf The regression model that provided the best fit to the remaining data is given below. The model was weighted by deepwater wavelength. Q’ = 0.035883 - 0.010479*P/1 + 0.005523*PI17 - 0.003424*P/2 + 0.001962*PI2” - 0.004667*PI4 + 0.000142* P14? + 0.000230* PIS” + 0.000000536*P/5* + 0.000068128*PI6 - 0.000000290*P/62 This model had a correlation coefficient of 0.976 (R2 = 0.952). The sum of squares of differences between predicted and measured dimensional overtopping rates was 0.887, which, for 98 data points, yielded an average difference of + 0.095 cfs/ft. As with all regression models presented in this report, this model is only valid for the range of conditions tested and is site specific for Revere Beach. The range of variables, both dimensional and nondimensional, used in this analysis is given in Table 14. Table 14 Minimum and Maximum Values for Parameters Used in Regression Analysis of Combined Data from Tasks A, B, and C | Parameter Min Max SWL, ft mlw 9.5000 16.6000 Wave height, ft 3.1000 12.6500 Freeboard between seawall and beach, ft 0.4000 11.8000 Cotan beach slope 10.7000 19.5000 Overtopping rate, cfs/ft 0.0000 1.3553 PH 0.0374 2.6818 Pla 0.3298 2.7188 PI4 10.7000 19.5000 PI5 0.8261 5.1875 PI6 33.0292 404.8744 0.0000 0.0256 Chapter 3 Research Tasks A, B, and C 43 44 4 Discussion of Research Tasks A, B, and C A major uncertainty in the physical model tests was the wave spectrum being tested. Wave information furnished for the storm profile consisted of a monochromatic wave height and wave period, obtained by refracting and diffracting a representative wave (peak period and significant wave height) of a wave spectrum. This representative wave, after shoaling to the approximate distance offshore modeled by the wave generator in the wave flume, then was used as the peak period and zeroth moment wave height to reproduce a new irregular wave spectrum. This would be accu- rate if the entire spectrum shoaled to the same extent as the representative wave. In reality, each frequency in the incident spectrum will shoal differ- ently, and an entirely new spectrum will exist after shoaling. Although we have the capability of dividing the incident spectrum into a number of bandwidths, transforming each bandwidth individually through numerical models REF/DIF and SBEACH, and then reassembling the transformed spectrum from the individual bandwidths, the procedure is time-consuming, not economically feasible, and other uncertainties in the prototype and physical model do not justify attempting such a level of precision. This uncertainty applied to Tasks A, B, C, and C+, and the net effect on over- topping rate caused by this approximation of the wave spectrum is unknown. With the 1978 profiles (Tasks A, C, and C+), there was considerable freeboard between the beach and seawall crest. Waves striking the sea- wall were forced into a vertical sheet of water and spray, frequently ex- ceeding the height of the seawall. Because the motion was nearly vertical, much of this water fell back on the seaward side of the seawall in the flume, but wind effects may cause more of the water to overtop the sea- wall in the prototype. Wind effects on overtopping rates in Task B are expected to be mini- mal. Wind has two effects on seawall overtopping rates: modification of wave runup on the beach, and blowing spray over the seawall. Modifica- tion of the wave runup has been calculated to have only a minor effect on Chapter 4 Discussion of Research Tasks A, B, and C overtopping!. Due to the low freeboard between the 1991 beach profiles and seawall crest elevations, waves overtopping the seawall tended to flow over the wall in a bore rather than be deflected vertically as in Tasks A, C, and C+. Because the water movement was horizontal rather than vertical, wind effects are not expected to be significant. Due to high reflection coefficients from the high seawall freeboard in models of the 1978 profile, wave energy reflected from the seawall re- mained in the wave flume and increased the total energy in the flume over time. Avoiding this effect would require that each test run be terminated before energy reflected from the structure could reach the wave generator and return to the structure. Each test then would be on the order of 2 min, after which the testing would be halted until the energy in the flume had dissipated. A series of short tests then would be used to ensure that the en- tire wave spectrum was represented. Again, this level of accuracy is prob- ably not justified, and would be time-consuming and expensive. The probable effect of this increased energy level in the flume is an increase in overtopping rates for the 1978 profile tests. Because of the low seawall freeboards and extended beach profiles compared to Tasks A, C, and C+, reflection coefficients for Task B were small and reflected wave energy was not a significant factor in the tests. Memorandum to Joan Pope entitled “Assessment of wind effects on wave overtopping of proposed Virginia Beach seawall,” 1987, from Donald T. Resio, Offshore and Coastal Technologies, Inc., Vicksburg, MS. Chapter 4 Discussion of Research Tasks A, B, and C 45 46 5 Revere Dike Study Park Dike, 1991 Profile Model construction Plans for the proposed park dike were received at CERC from NED, and are reproduced in part in Figure 13. The plans propose narrowing Revere Boulevard and adjacent sidewalks from the current 80 ft to 46 ft and building a mound of random fill covered with a 12-in. layer of top- soil. Design of the mound specified a crest elevation of +27.5 ft mlw (equivalent to 23.0 ft NGVD as seen in Figure 13, where NGVD is Na- tional Geodetic Vertical Datum), a toe wall along the seaward toe at eleva- tion +21.5 ft mlw, a width of approximately 110 ft from toe wall to crest yielding a slope of approximately 1:18.3, and a slope of 1:2.5 on the land- ward side of the dike. Within the dike is a rubble core to provide protec- tion in case the topsoil and random fill are eroded away during a storm. In the model, a toe wall was constructed to an elevation of +21.5 ft mlw at a distance of 46 ft shoreward of the seawall as specified in the park dike plans. The dike was constructed with plywood extending from the crest of the toe wall to the crest of the dike, with the area shoreward of the dike crest sealed to retain any overtopping (Figure 14). All tests were conducted by first filling the area between the seawall and toe of the park dike with water to allow maximum wave energy to reach the dike. Seawall elevations of +20.9 and +21.3 ft mlw are found in the reach represented by Profile 2; both seawall elevations were tested. Three park dike crest elevations were tested. The highest crest eleva- tion was the proposed elevation of +27.5 ft mlw, the lowest elevation was set at +24.0 ft for a seaward slope of about 0.014, and the third elevation was approximately midway at +25.6 ft mlw. The model was tested with the 1991 post-storm beach profile which had been constructed during Task B, above, for Profile 2. The beach pro- file had a beach elevation at the seawall of +20.5 ft mlw. Additional tests Chapter 5 Revere Dike Study "aq 3LV1d SNOILISS ONY NV1d SIG WYVd HOV3SE 3Y3A3yY AGNLS NOILONG|Y 39vVWVG GOOls saiuvingidl GNV YaAIN snonvs ABV AYHLIA SU33NION3 JO SdYOD NOISIAIG GNVIDN3 MIN AWYW 3HL JO IN3WLYVd30 Creea GApaRe 200459 OF ,7 + Wig Pues 1 all 420lo4y 42029 SSS SEI IF Pian PWNZ 0M Cuisclry. AEM 2p, O'Z/ 17 /8R9/ Bel sadoms/i4o ubiseg -— ¥LOM ap/s S31VOS dIHKdved 200g 3130247 ayIp ysed Ie 17 ‘Nemeas wWys0d -4si¥7— bL/ 77 fomy/om 45/xX7 PVG [207G 212A2Y YOM apis 00 sez Oe BST ar (02 #SIKzr $ ‘S, G1. (eee Creyo CF) 0L/ 17 EM C1111 04 Opy- Ald Bompooe EC HOM BPS, B DOW FF PPSsedory YIM FO fp M7 ,0v, VZIWOH.ssy95 > =,1 LUBA 9-9 NOILI3S Nera Buruiopay 2002, pial sa0d 4s/n2 array / a | Iy Yopuos prf2edwog 280g jaw 2/ | 02 79! pesodod au} 10} suejd adAjojo/d Jo uolod vs ive, 03, -V LY3aA “ZINOH NOILI3SS | ‘avs Wg saz asedu! Per2w3s 2g 04 Worm ves 2022 fS/x7 LOM Ess togas ByevoiveD Bon yjom [2 POA 229G BID ADY 4912 hom poos 4sixz fompoos NS piospsrne yb Ol =,'3TVIS 8-98 NOILD3SS 247 CAP Pepses ‘7195 Peg AInfs 'X27 Can SIO IS. Purosb b6UUSIXF < Mes (egples Pafredtiuco> P2eP322eS (12SAL 2 OSOMIAL0C4F D104 S BL Ouwprem (20019 1a/ “napiiey Po -predue \ Cee 77 \ Prepress rsed2g 2 €} eunbi4 DAV VOPIC £0 >PF S a é S ~ $ co 1 ary Ucar J? PUPA NR = Chapter 5 Revere Dike Study +20.9 or +24.0 ft +21.5 ft +21.3 ft Park Dike Sea- wall Concrete Slope in Wave Flume All elevations in feet mlw Figure 14. Cross section of model park dike were conducted with beach elevations at the seawall of +14.4, +15.4, and +19.9 ft mlw to test the effects on overtopping rates of an eroded beach profile. Test conditions and results Tests were conducted for a prototype wave with H,,,, = 12.7 ft, the larg- est wave height of the design storm. Prototype wave periods tested were T,, = 15.9 sec and 13.0 sec, corresponding to the peak periods of the SPN and 1978 storm, respectively. Water depth was +16.6 ft mlw, the greatest depth of the design storm. The beach profile tested was the 1991 post- storm profile with a maximum beach elevation of +20.5 ft mlw. With these test conditions, there was no measurable overtopping with the low- est crest elevation tested (+24.0 ft mlw); therefore, higher crest elevations were not tested under design storm conditions. A series of tests under less severe conditions had been planned, but were not conducted because of the lack of overtopping under the most extreme conditions of the SPN. Test conditions and measured overtopping rates are listed in Table 15. Tests 1 and 2 were conducted with a seawall crest elevation of +21.3 ft mlw and peak wave periods of 15.9 and 13.0 sec, respectively. Tests 3 and 4 were conducted under the same wave conditions, but with the sea- wall crest lowered to +20.9 ft mlw. Minor overtopping was observed in all four tests, but the overtopping quantities were not sufficient to mea- sure. Qualitatively, higher overtopping rates were observed with the higher seawall crest elevation and higher wave period. The sheet metal beach in front of the seawall was removed for Tests 5 and 6 to determine effects of beach erosion on overtopping rates. With a beach elevation of +15.4 ft mlw, overtopping rates increased considerably, although the overtopping was still not measurable at the 13.0-sec peak wave period. For Test 7, the swl was increased by 1 ft to +17.6 ft mlw to simulate possible sea level rise during the life of the structure. The additional foot 48 Chapter 5 Revere Dike Study Table 15 Test Conditions and Overtopping Rates for Park Dike with 1991 Profile SWL Run No. | ft mlw 1 16.6 16.6 — Seawall Dike Crest Crest Beach Over- Elev. Elev. Elev. topping ft mlw ft mlw ft mlw cfs/ft mat 24.0 20.5 ' Qvertopping too low to be measured. Overtopping of seawall, not park dike. EEE of depth greatly increased the overtopping rates to nearly 0.007 cfs/ft (prototype). Test 8 increased the seawall crest elevation by 5 ft to +26.9 ft mlw to determine if an increase in seawall crest would prevent overtopping with- out the expense of the park dike. Overtopping was measured directly be- hind the seawall and averaged 0.008 cfs/ft. The additional foot of depth used in Test 7 was not used in Test 8. Tests 9 through 11 brought the swl back up to +17.6 ft mlw. Tests 9 and 10 raised the crest of the park dike to +25.7 ft mlw and used beach elevations of +14.4 and +19.9 ft mlw, respectively. Test 11 raised the park dike crest elevation to +27.5 ft mlw, returned the seawall crest eleva- tion to +21.3, and replaced the sheet metal slope in front of the seawall to a beach elevation of +20.6 ft mlw. The park dike was overtopped in all three tests. Chapter 5 Revere Dike Study 49 Rubble-Mound Dike, 1991 Profile Model construction Modeling of the rubble-mound dike assumed the prototype would be constructed with an impermeable core covered by an underlayer and two layers of armor stone. The model rubble-mound dike was constructed of a piece of plywood for the impermeable core with crushed gravel retained by a No. 6 sieve glued to the board to simulate the underlayer. Crushed gravel passing a 3/4-in. sieve and retained by a 5/8-in. sieve was used for the armor stone. Average weight of the armor stones in the model was 0.022 lb (672 lb prototype). The toe wall used for the park dike was used again for the rubble- mound dike but repositioned further back from the seawall on the assump- tion the mound would be built on the west side of the existing Revere Boulevard. Crest height of the impermeable core was set at +25.4 ft mlw. The rubble mound was constructed to the dike crest, and the area behind the crest was sealed to retain the overtopping (Figure 15). +25.4 4215 ft +21.3 ft Rubble-Mound Sea- Dike wall Concrete Slope in Wave Flume All elevations in feet mlw Figure 15. Cross section of rubble-mound dike A seawall crest elevation of +21.3 ft mlw was used for all tests because it was observed in the park dike study that this crest elevation produced more overtopping than the lower crest elevation. Test conditions and results Generally, the same test conditions used in the park dike study were used for the rubble-mound dike. Still-water level was at +16.6 ft mlw, and beach elevation was set at +20.5 ft mlw. Test conditions and overtop- ping rates are listed in Table 16. Tests 12 and 13 used a peak wave period of 13 sec and H,,,,’s of 10.0 and 12.7 ft, respectively. There was no overtopping in either test. 50 : Chapter 5 Revere Dike Study Table 16 Test Conditions and Overtopping Rates for Rubble-mound Dike with 1991 Profile Seawall i Wave Wave Crest Over- Height Period Elev. : 0 topping ft sec ft mlw cfs/ft 10.0 5 us H u Overtopping too low to be measured. Tests 14 and 15 used a peak wave period of 15.9 sec and H,,,,’s of 12.7 and 11.0 ft, respectively. Again there was no overtopping. Although there was no overtopping in Tests 12 through 15, there was some splashing over the rubble mound. The quantity of splashing was too small to measure. Wave runup approached the crest of the mound without flowing over the crest. For Test 16, the swl was raised 1 ft to +17.6 ft, simulating possible sea level rise. Peak period was 15.9 sec and wave height was 12.7 ft. The high-water level produced an overtopping rate of 0.001 cfs/ft. For Tests 17 and 18, the beach in front of the seawall was removed, leaving a beach elevation of +15.4 ft mlw. Still-water level was returned to +16.6 ft mlw for Test 17 and +17.6 ft for Test 18. Overtopping was observed in both tests but was not sufficient to measure at the lower water level (Test 17). Overtopping rate for Test 18 was 0.0005 cfs/ft. For Tests 19 and 20, the beach was partially restored to an elevation of +17.7 ft mlw. The still-water level was set at +16.6 ft mlw for Test 19 and raised to +17.6 ft for Test 20. Overtopping was observed in both tests, but was not measurable. There was no armor instability in any of the tests in this test series. Chapter 5 Revere Dike Study 51 52 Park Dike, 1978 Profile Model construction The park dike model was constructed identically to the model with the 1991 profile, except that tests with the 1978 bathymetry were conducted in the 18-in. flume rather than the 3-ft flume. Crest elevation of the park dike was +24.0 ft mlw. Test conditions and results Table 17 lists conditions used to test the park dike. The swl’s selected were the highest water levels predicted for the SPN and the highest swl observed during the 1978 storm. Similarly, wave periods were maximum wave periods for the SPN and the 1978 storm. Wave heights were se- lected as maximum wave height of the SPN plus lower wave heights to provide a range of overtopping values. See Table 4, Profile 2, for SPN. conditions, and Table 2 for conditions in the 1978 storm. The wave generator in the 18-in. flume was unable to produce the maxi- mum wave conditions selected for testing. For test conditions where the wave generator was unable to produce the desired wave spectrum, the H,,, was incrementally decreased by 10 percent until conditions were within the limits of the wave generator. Water depths and wave periods were not changed. Table 17 lists wave heights that were used in the test series. Tests 21 through 33 in Table 17 were tested with the same geometry of beach, seawall, and dike elevations. For the remaining tests (34 though 45), the swl was kept constant at +16.5 ft mlw and the same dike elevation of +24.0 ft mlw was maintained. Four of the test conditions in Table 17 (two wave heights at each of the two wave periods) were selected for each of the three remaining sets of tests. Tests 34 through 37 measured overtop- ping rates with the beach elevation raised to +13.1 ft mlw, and Tests 38 through 41 further increased the beach elevation to +16.8 ft mlw. For Tests 42 through 45, the beach was returned to the 1978 profile elevation of +9.3 ft mlw and the seawall crest elevation was reduced to +18.5 ft mlw (approximate elevation of Revere Boulevard in the vicinity of Pro- file 2). Tests 42 through 45 were meant to simulate conditions if the sea- wall were to fail. Testing the park dike with the 1978 bathymetry produced very large overtopping quantities. At swl = +16.5 ft mlw and 7, = 15.9 sec, the smallest wave heights tested (H,,,, = 8.9 ft, Test 21) caused sheets of water to completely overtop the structure, with some spray passing over the en- tire park dike while still in the air. Vertical spray exceeded the height of the flume, and a board was placed on top of the flume to retain the spray. These extreme cases of overtopping occurred when groups of large waves Chapter 5 Revere Dike Study | Table 17 Test Conditions and Overtopping Rates for Park Dike with 1978 | Profile Seawall Dike Wave Wave Crest Crest Beach Over- SWL Height Period Elev. Elev. Elev. topping Run No. | ft mlw ft sec ft mlw ft mlw ft mlw cfs/ft 21 16.6 8.9 15.9 21.3 24.0 0.0855 | 21.3 24.0 0.0960 23 15.9 21.3 24.0 0.0844 24 16.6 11.4 13.0 21.3 0.0836 11.0 De) ()] ie) N 8.4 7.2 o no ) ro) aa 55 © 13.0 13.0 15.9 13.0 C1 1) 1) PD PG 1 Co | w | |w 21.3 21.3 21.3 21.3 21.3 21.3 9.3 9.3 9.3 9.3 9.3 9.3 0.0713 0.0642 0.0015 0.0041 0.0002 0.0017 31 14.8 13.0 21.3 24.0 0.0026 “i [me en jas lea | a 33 17.6 8.9 15.9 24.0 9.3 0.2574 34 16.6 8.9 15.9 24.0 13.1 0.0220 Ea ce 7a ees pases [es oo a Ie =n Co [ia eo SRS) 24.0 13.1 0.0207 8.9 15.9 21.3 [240 = fies | 0.0121 10.0 15.9 21.3 24.0 16.8 0.0162 13.0 21.3 24.0 16.8 0.0058 24.0 16.8 0.0000 24.0 9.3 0.0719 16.6 10.0 15.9 18.5 24.0 9.3 0.0769 44 16.6 11.4 13.0 18.5 24.0 9.3 0.0612 45 [16.6 — 10.0 13.0 18.5 24.0 9.3 0.0644 prevented rundown on the slope and produced a hydraulic head between the park dike and the seawall. Test 22 increased the wave height to 11 ft and produced greater over- topping. Test 23 used the same conditions as Tests 21 and 22, except for a wave height of 10.0 ft. Although the wave height in Test 23 was 1.1 ft greater than in Test 21, the measured overtopping was less by about | per- cent. This slight discrepancy could be caused by the random nature of the Chapter 5 Revere Dike Study 53 54 wave trains being used or by inaccuracies in the collection and measure- ment. However, Tests 27 through 29 differed only in wave height, and Test 27 with a wave height of 9.6 ft had a low overtopping rate relative to Tests 28 and 29. Similarly, Tests 30 through 32 differed only in wave height, and Test 30 with a wave height of 9.9 ft had a low overtopping rate relative to Tests 31 and 32. In each of these sets of tests, wave heights around 9 to 10 ft were seen to produce surprisingly low overtopping. This trend of low overtopping rates was observed only with a beach elevation of +9.3 ft mlw and was not observed in Tests 34 through 41, which used a higher beach elevation, or in Tests 42 through 45, which used a lower sea- wall elevation. As expected, reducing the swl to +14.8 ft mlw in Tests 28 through 32 greatly reduced overtopping rates, while increasing the swl to +17.6 ft in Test 33 nearly inundated the structure. Raising the beach elevation in front of the seawall to +13.1 ft mlw in Tests 34 through 37 decreased overtopping rates, and further increasing the beach elevation to +16.8 ft in Tests 38 through 41 further decreased overtopping rates. The only test of the park dike with the 1978 profile that did not produce overtopping was Test 41 with the beach elevation at +16.8 ft mlw. Tests 42 through 45 returned the beach profile to the conditions of the 1978 survey (+9.3 ft mlw) and reduced the seawall elevation 2.8 ft to +18.5 ft mlw. Overtopping rates were less than under the same conditions but with the seawall intact (Tests 21, 23, 24, and 26) for Tests 42 through 44, and showed little change in Test 45. This was consistent with the find- ing reported above in the tests of the park dike with the 1991 profile; i.e., higher overtopping rates were obtained with the higher seawall elevation. Rubble-Mound Dike, 1978 Profile Model construction The rubble-mound dike model was constructed in the same manner as the 1991 profile, but a smaller armor stone was used. Although specific tests for stability were not conducted, there was no movement of armor stone observed on tests with the 1991 profile. The armor stone was there- fore reduced to crushed gravel passing a 5/8-in. sieve and retained by a 1/2-in. sieve. The model armor stone had an average weight of 0.011 Ib per stone (336 lb prototype). Chapter 5 Revere Dike Study Test conditions and results Three sets of four tests each were conducted with each set consisting of one wave height at each of two wave periods at each of two swl’s. Test conditions were the highest obtainable wave height at each of the wave periods and swl’s listed in Table 17, with the exception of swl = +17.6 ft mlw, which exceeds the design storm conditions and was not tested in this series. Tests 46 through 49 were conducted with the beach elevation at +9.3 ft mlw (1978 survey), Tests 50 through 53 repeated the wave condi- tions but with the beach elevation raised to +14.3 ft mlw, and Tests 54 through 57 raised the beach elevation to +16.7 ft mlw. Test conditions, measured overtopping rates, and number of armor stones displaced are listed in Table 18. Although the rubble-mound dike never approached a failure condition, with failure defined as having the underlayer exposed, armor stones were displaced during several of the tests. Displaced armor stones were re- placed on the structure only after each set of four tests. Tests 46 through 49 (beach elevation as measured in the 1978 survey) all produced overtopping. Some armor stones were displaced, with - Table 18 Test Conditions, Overtopping Rates, and Armor Stone Displacement for Rubble-mound Dike with 1978 Profile Armor Stone Displacement Wave Over- Run SWL Height 0 b : topping No. ft mlw | ft cts/ft Seaward Shoreward res _|io [ise [ere lesa [os locos [is | Cee ae pe oe ee eee ee — 0.0024 Chapter 5 Revere Dike Study = 56 15 stones moved to in front of the toe wall and 18 stones carried over the crest of the dike during Test 46, 9 stones displaced to in front of the toe wall in Test 47, and 3 stones displaced to seaward of the toe wall in Test 49. Displaced stones were not replaced until after Test 49. For Tests 50 through 53, the beach elevation in front of the seawall was raised to +14.3 ft mlw. There was no overtopping in Test 50 and only a very small and unmeasurable overtopping from one wave in Test 51. Neither Test 50 nor 51 had any armor stones displaced. Tests 52 and 53 had measurable overtopping, with six armor stones displaced seaward and eight armor stones displaced shoreward during Test 52, and two armor stones displaced seaward in Test 53. Tests 54 through 57 raised the beach elevation in front of the seawall to +16.7 ft mlw. Minor overtopping was observed during Tests 54 and 55 with two armor stones displaced seaward in Test 54. There was no over- topping and no armor stone displacement in Tests 56 and 57. Armor Unit Stability Stability of armor units on the rubble-mound dike was not specifically tested, but the following information may be of value for design purposes. As reported above, there was no armor stone displacement using stones with an average weight of 0.022 lb and bathymetry from the 1991 survey. Armor stone displacement during tests with the 1978 bathymetry and armor stones averaging 0.011 Ib are given in Table 18. Armor stones used in the tests were a crushed dolomite with a unit weight of 165 pcf. Based on relationships defined by Froude’s model law (see Chapter 2, “Test Fa- cility”), the following transference equation is derived to determine proto- type stone weights. (WY) fr. P [Sd - 17 m m |_m Pp i Gp | IG =i P Pp m where W,, = weight of an individual stone, Ib subscripts m,p = model and prototype values, respectively Yq = specific weight of an individual stone, pcf Lyp/Lp = linear scale of the model Chapter 5 Revere Dike Study S, = specific gravity of an individual stone relative to the water in which the breakwater is constructed, Ey, Sern Yw = specific weight of water, pcf Assuming a specific weight of seawater of 64.0 pcf and fresh water of 62.4 pcf, assuming a specific weight of 165 pcf for both model and proto- type stone, and using a model scale of 1:30, average weights of armor stone used in the models correspond to average prototype weights of 672 |b and 336 lb for tests conducted with the 1991 profile and 1978 pro- file, respectively. Chapter 5 Revere Dike Study 57 58 6 Revere Dike Discussion The proposed park dike with the crest lowered to provide only a 1.5-ft rise from toe wall to crest was found to be sufficient to prevent nearly all overtopping during the design storm event using the post-storm 1991 beach profile. Revere Boulevard, of course, would be completely flooded. In the model, waves overtopping the seawall and crossing Revere Boule- vard would flow part way up the park dike in a solid sheet of water across the width of the flume. As the runup decreased, the sheet of water would be reduced to a few “fingers” or thin streams of water that flowed much further up the slope of the dike. Under the most severe conditions of the SPN, most sheet flow did not extend more than one-third to one-half the distance to the crest of the dike before separating into a few “fingers.” All overtopping observed with the park dike under design storm condi- tions with the 1991 bathymetry occurred when one of the “fingers” reached the crest of the dike. At no time did the solid sheet of runup reach the crest. It is anticipated that prototype runup on a park dike cov- ered with vegetation and paths would be less than observed in the wave flume. If the profile in front of the seawall is maintained at a bathymetry similar to the 1991 survey, the park dike with a crest elevation of +24.0 ft mlw should be adequate to prevent nearly all overtopping during the de- sign storm event. It should be recognized, however, that in any random sea event there is a possibility of an event occurring that exceeds the con- ditions tested in the physical model. Decreasing the elevation of the seawall decreased the rate of overtop- ping over the park dike. With the toe wall maintained at a constant eleva- tion, decreasing the freeboard of the seawall increased the freeboard of the toe wall over the seawall by the same amount, increasing the effective- ness of the toe wall. Although there was more overtopping of the lower seawall, the increased effectiveness of the toe wall resulted in less water overtopping the dike. Tests conducted to determine effects of a failure of the seawall to the elevation of the roadway found that overtopping rates were lower than with the seawall in place. Wave breaking occurred either on the slope in front of the seawall or across the seawall onto Revere Boulevard. By the time wave action crossed the toe wall onto the park dike, most of the turbulence had dissi- pated, and flow on the dike appeared to be predominantly laminar. If the Chapter 6 Revere Dike Discussion park dike is covered with dirt and vegetation, scour can be expected dur- ing peak levels of the design hydrograph. Extensive scouring is not ex- pected, however, due to the dissipation of the turbulence and the short duration of the hydrograph peaks. It is doubtful that the rockfill inside the park dike specified in the plans is necessary. A small rubble mound with an impervious core was found to be effec- tive in preventing overtopping when tested with the 1991 profile. The roughness of the stone structure quickly halted the runup, and a much smaller structure than the park dike was found to be sufficient. With a crest elevation of +25.4 ft mlw, there was no overtopping during design storm conditions, with the exception of a minor quantity of splashing. However, runup was observed to approach the mound’s crest, and overtop- ping would have occurred at a lower mound crest elevation. Overtopping was observed on tests conducted to simulate beach erosion in front of the seawall, increased swl from sea level rise, or on tests with the 1978 profile. Displacement of armor stones occurred with stones averaging 336 lb (prototype) during tests with the 1978 profile. It should be noted that armor stone displacement is common with new construction, and typically decreases as the stones become seated by wave action. Although the num- ber of armor units displaced decreased during each successive set of tests with the 1978 profile and the rubble-mound dike, the amount of wave ac- tion on the dike was less during each successive set of tests due to in- creases in the beach elevation. Because wave action on the dike with the 1991 profile was less than with the 1978 profile, and assuming that stones would be seated during storms of less severity than the design event, the 336-Ib stones are probably sufficient if the 1991 beach profile is maintained. If the beach profile returns to a bathymetry similar to the 1978 profile, both park dike and rubble-mound dike will be overtopped during the de- sign storm event and under less extreme conditions. Because sea condi- tions varied for the various tests conducted under this research effort, it is difficult to compare overtopping rates for the different profiles and struc- ture options at Revere Beach. However, the following comparison, based on conditions at the peak of the SPN, may be instructive. Table 19 lists several tests of Profile 2 tested at the peak of the SPN with an swl of +16.6 ft mlw and a wave period of 15.9 sec. Both 1978 and 1991 profiles are included, as are both park dike and rubble-mound dike conditions, as well as overtopping rates without a dike. By far the highest overtopping rate was found with the 1978 profile and no dike. With the addition of the beach fill (1991 profile), the over- topping rate was reduced by about 65 percent even with a wave height that was half again as high as that conducted on the 1978 profile. The addition of either the park dike or rubble-mound dike, with the 1991 profile, re- duced the overtopping to nearly zero. The physical model did not include erosion of the 1991 beach profile during the SPN storm. Overtopping rates may be expected to range between those measured with the 1991 and Chapter 6 Revere Dike Discussion she) 60 Table 19 Comparison of Overtopping Rates at Peak of SPN Year of Survey Dike 1978 None 1991 None 1991 Park 1991 Rubble-Mound 1 1978 7 18 1978 ' Overtopping rate too small to be measured. (L_— Table in which data are listed 12 8 Wave Period 15 Ll 1978 profiles because erosion of the 1991 beach profile during the SPN storm can be expected. If the beach erodes back to the 1978 profile, overtopping of either dike will occur. The last two entries in Table 19 give overtopping rates for the park dike and rubble-mound dike with the 1978 profile. The dikes greatly reduced the overtopping rate compared to conditions without any dike, al- though the overtopping rates with either dike may still be unacceptable if the beach is eroded to the 1978 profile. As a qualitative reference, Fukuda, Uno, and Irie (1974) measured and filmed waves overtopping a seawall fronted by a concrete revetment dur- ing severe storms. The films were then viewed by a panel of coastal ex- perts who estimated the degree of danger posed by the overtopping. Averaging the results of the panel, it was determined that at a location 10 ft behind the structure, overtopping rates greater than 0.0002 cfs/ft would prohibit a vehicle from driving past at high speed, damage to a house could be expected at an overtopping rate of 0.0007 cfs/ft, and over- topping rates greater than 0.002 cfs/ft would be dangerous for a walking person. These overtopping rates assumed an average over several hundred waves and could be increased by a factor of 10 for a location 30 ft behind the structure. For protection of a relatively densely populated coastal area, Goda (1985) reports an overtopping rate of 0.1 cfs/ft as an adopted guideline in port areas in Japan. These overtopping rates assume an aver- age over several hundred waves (Goda 1985). Based on the overtopping rates given above from Fukuda, Uno, and Irie (1974), overtopping rates with either dike may be hazardous if the beach erodes to the 1978 condition. However, conditions listed in Table 19 occurred only at the peak of the SPN hydrograph and the dikes eliminated 95 to 98 percent of the overtopping compared to the 1978 condition. Lower swl’s caused substantially less overtopping (see Tables 17 and 18). Chapter 6 Revere Dike Discussion References Fukuda, N., Uno, T., and Irie, I. (1974). “Field observations of wave overtopping of wave absorbing revetment,” Coastal Engineering in Japan 17. Goda, Y. (1985). Random seas and the design of maritime structures. University of Tokyo Press, Tokyo, Japan. Hasselmann, K., Barnett, T. P., Bouws, E., Carlso, H., Cartwright, D. C., Enke, K., Ewing, J., Gienapp, H., Hasselmann, D. E., Sell, W., and Walden, H. (1973). “Measurements of wind-wave growth and swell decay during the joint sea wave project (JONSWAP),” Deutshes Hydrographisches Institut, Hamburg, Germany. Hughes, S. A. (1984). “The TMA shallow-water spectrum description and applications,” Technical Report CERC-84-7, U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS. Smith, W. G., Rosati, J. D., Bratos, S. M., and McCormick, J. (1994). “Revere Beach and Point of Pines, Massachusetts, Shore Front Study,’ Miscellaneous Paper CERC-94-1, U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS. Stevens, J. C., Bardsley, C. E., Lane, E. W., and Straub, L. G. (1942). “Hydraulic models,” in Manuals on Engineering Practice No. 25, American Society of Civil Engineers, New York. References 2 61 Hl ote: phys hier ea io f t Pets ENVY dA: PROC Li al ehabia e hela Ca gocaaliarat i) eee Bt a Whee SRE iene com op RE Menno camber earulid Lae, = Fea 2 ee) f Gal fr, ¢ fey am , , t er 3 PD y j oi in : : me J My) bs F yeah ; EF aes a | 5 IG an y y os ih A a ; aon F, - m Pee Taal m i tae F Fe f ele i BON ty Lf iJ j , ’ is yi ; Men } , " g f Woke te ; i 4% Teele. 4 = Nat 4 4 : ; ' ey i & ; Pant fl “yi Lae BY yee i= Bray i Far euane 5 i Y b i Gah oe é ia ERIN steko z en Hea ie a rene ‘ ‘ 7° 5 eee : i y X f REPORT DOCUMENTATION PAGE AR ono es Public reporting burden for this collection ofinformation is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining || the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-4302, and to the Office of Management and Budget, Paperwork Reduction Project (0704-0188), Washington, DC 20503. la. TITLE AND SUBTITLE 5. FUNDING NUMBERS Physical Model Study of Revere Beach, Massachusetts . AUTHOR(S) Donald L. Ward PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION U.S. Army Engineer Waterways Experiment Station BE OnT NUMBER 3909 Halls Ferry Road, Vicksburg, MS 39180-6199 Technical Report CERC-95-2 U.S. Army Engineer Division, New England AGENCY REPORT NUMBER 424 Trapelo Road Waltham, MA 02254-9149 11. SUPPLEMENTARY NOTES Available from National Technical Information Service, 5285 Port Royal Road, Springfield, VA 22161. }12a. DISTRIBUTION/AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORING/MONITORING | |13. ABSTRACT (Maximum 200 words) Physical model studies were conducted in 2-D wave flumes to determine overtopping rates for existing and proposed structures along Revere Beach, Massachusetts, during design storm events. Results of the physical model studies were used in regression |analysis to develop simple nondimensional equations relating overtopping rates to incident wave conditions and structure design |along various reaches of the beach. Total volume overtopping in the physical model was compared to prototype data for a known storm event to verify the model. Overtopping rates were then determined for the existing seawall using bathymetries surveyed both |before and after a 1991 beachfill project. The effectiveness of a proposed dike shoreward of the seawall was also measured in the physical model. | Approved for public release; distribution is unlimited. 14. SUBJECT TERMS Coastal flooding Regression analysis Dike Revere Beach, Massachusetts Overtopping Seawall 15. NUMBER OF PAGES 69 16. PRICE CODE SECURITY CLASSIFICATION |19. SECURITY CLASSIFICATION |20. LIMITATION OF ABSTRACT OF THIS PAGE OF ABSTRACT UNCLASSIFIED NSN 7540-01-280-5500 Standard Form 298 (Rev. 2-89) Prescribed by ANSI Std. Z39-18 298-102 SECURITY CLASSIFICATION |18. OF REPORT UNCLASSIFIED ene Fh nea vi a i vt srt oan A SRK Voie abel) Selb Wyle daipai og es {ire Ge mers i il 2 at OY Rol DEPARTMENT OF THE ARMY WATERWAYS EXPERIMENT STATION, CORPS OF ENGINEERS 3909 HALLS FERRY ROAD SFECTAL VICKSBURG, MISSISSIPPI 39180-6199 FOURTH CLASS ——— u.5. POSTAGE FALE Official Business VICKSBURG, AS PERMIT HO. 85 ea7/Liez/ i WOODS HOLE OCEANOGRAPHIC INSTITUTION COASTAL RESEARCH CENTER WOODS HOLE FA 02543-1525