UFL / COEL-89 / 008
PERFORMANCE PREDICTION OF BEACH NOURISHMENT PROJECTS
W. Samuel Phlegar III
PERFORMANCE PREDICTION OF BEACH NOURISHMENT PROJECTS
W. SAMUEL PHLEGAR III
A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF MASTER OF SCIENCE
UNIVERSITY OF FLORIDA
To begin, I would like to sincerely thank my advisor and supervisory committee chairman, Dr. Robert G. Dean, for his insight, direction, guidance, and support throughout this project. I consider it a true honor to have worked under his patient leadership. I would also like to thank Dr. Max Sheppard and Dr. Ashish J. Mehta for their contribution during the important editing and finalizing stages.
The Florida Sea Grant Program and the National Park Service provided the sponsorship upon which this research is based and this support is greatly appreciated.
Thanks go to Lillien Pieter for the drafting of many of the figures and for her patience in times of ultimate crisis. The personal conversations with Kim Beachler of Coastal Planning and Engineering and detailed reports found in the excellent Coastal Engineering Archives provided important background information. Also, the work of Don Stauble in the monitoring aspects of coastal projects and the summary of past projects is acknowledged and appreciated.
Many thanks, and my future, go to Miss Kathy L. Hammock, soon to be Mrs. Kathy Phlegar, for her smiles, patience, love, and finally for not deciding to purchase a handgun and end her "headache" during and after the thesis preparation days. I must thank my parents, Bo and Joanne Phlegar, most of all for their ultimate support throughout my life and for giving me the wonderful ability to smile and laugh through most situations. If we couldn't laugh, we would all go insane.
Finally, thanks to those who know who they are: Jimmy B., Wesley, Victim, Easton, Issac N., Gusty, Aggie Douglas, Peener, Therapy Too, Mr. B., Jeffy, Billy, and B.J.
TABLE OF CONTENTS
ACKNOWLEDGEMENTS ...... LIST OF FIGURES .........
LIST OF TABLES .........
1 INTRODUCTION ........
1.1 Purpose of Study ......
1.2 Project Selection ...... 2 APPROACH .............
2.1 Review of Methodologies 2.2 Relevant Parameters ..
2.2.1 Introduction ..
2.2.2 Wave Characteristics 2.2.3 Background Erosion
2.2.4 Local Conditions 3 METHODOLOGY ....
3.1 Governing Equations ..
3.1.1 Continuity Equation
3.1.2 Dynamic Equation 3.1.3 Combined Equation
3.2 Analytic Solution .
3.3 Numerical Solution .
.. . . ii
. . .
. . .
. . .
. . .
. . .
3.4 Wave Parameters . .
3.5 Remaining Parameter Summary .
3.5.1 Wave Refraction . .
3.5.2 Sediment Size.. . .
3.6 Model Summary . .
4 PROJECT APPLICATION AND RESULTS
4.1 Introduction . . ..
4.2 Application and Results for Each Location .
4.2.1 Delray Beach . .
4.2.2 Cape Canaveral . .
4.2.3 Indialantic Beach . .
4.2.4 Jupiter Island . .
4.2.5 Ft. Pierce . .
4.2.6 Hillsboro . .
4.2.7 Pompano Beach . .
4.2.8 Hollywood/Hallandale .
4.2.9 Treasure Island . .
4.2.10 Captiva Island.. . .
5 CONCLUSIONS . . .
5.1 Summary of Investigation . .
5.1.1 Prediction Analysis . .
5.1.2 Parameter Sensitivity .
5.1.3 Application Limitations .
5.1.4 Nourishment Monitoring Importance
5.2 Recommendations for Future Work BIBLIOGRAPHY . . .
BIOGRAPHICAL SKETCH . .
LIST OF FIGURES
2.1 Location of coastal data network stations maintained by the Coastal and
Oceanographic Engineering Department at the University of Florida. 7
2.2 Three possible boundary condition applications: 1) open coastline. 2)
structure on beach with minimal net longshore transport. 3) inlet with
a dominant transport direction . . 11
3.1 Profile displacement in response to accretion or erosion . 15 3.2 Sketch showing shoreline orientation and parameter explanation 17
3.3 Shoreline evolution for initially rectangular fill, I = 4miles, H = 2ft.,
Width = 100ft. (from Dean, 1988) . . 19
3.4 Proportion, M(t), of fill remaining within original project limits. ..... 19 3.5 Schematic representation of shoreline for numerical application. .... 21 3.6 Plot of sediment diameter and the k value (after Dean, 1988) 27 4.1 Location of beach nourishment projects to be studied. . 30
4.2 Delray Beach area and project locations . . 34
4.3 Delray Beach area background erosion: shoreline change plotted at DNR
monuments for 1945 -1970 . . . 35
4.4 Delray Beach: model prediction and survey volumes . 36
4.5 Delray Beach: parameter variance plot; contours of variance calculations
X 1010. . . . . 37
4.6 Delray Beach: k parameter variance with constant (h. + B) =24.0 38 4.7 Delray Beach: refraction effects on prediction of evolution 39
4.8 Cape Canaveral: project and borrow area locations. (after Stauble,
1985) . . . . 42
4.9 Cape Canaveral: survey volumes and resulting model volumetric changes
for two cases of varying transport rate . . 43
4.10 Indialantic Beach: project location and borrow site. (after Stauble, 1985) 46
4.11 Indialantic Beach: bar diagram of planform and grid cell network after
fill ......... .......... ......................... ....47
4.12 Indialantic Beach: model and analytical prediction showing survey volum es. . . . . 48
4.13 Indialantic Project: erosion rate sensitivity, +2 to 5 ft/yr. 49
4.14 Jupiter Island: project locations for multiple nourishments. (Modified
from Aubrey, 1988). -. . . 52
4.15 Jupiter Island: model volumes and indication of possible survey variance. 53 4.16 Jupiter Island: erosion rate sensitivity, 0 to 8 ft/yr range. . 54 4.17 Ft. Pierce: project and inlet location. . . 58
4.18 Background erosion rates for St. Lucie County shorelines. 59 4.19 Ft. Pierce: model predicted volumetric changes. . 60
4.20 Hillsboro Beach: nourishment project area . 63
4.21 Hillsboro Beach: survey volumes and model prediction . 64 4.22 Pompano Beach: project locations for 1970 and 1983 beach fills 67
4.23 Hillsboro Inlet: jetty construction history and impoundment basin location ........ ................. ..................... 68
4.24 Pompano Beach: model predictions and survey results . 69 4.25 Hollywood/Hallandale: project and borrow area locations. 73 4.26 Hollywood/Hallandale: survey, analytical, and model volume results 74
4.27 Hollywood/Hallandale: sensitivity to background erosion rate, 2 ft/yr
and 4 ft/yr. . . . . 75
4.28 Treasure Island: nourishment project areas and proximity to Blind Pass
and John's Pass. . . . 78
4.29 Treasure Island: survey volumes and model prediction through a two
year period. . . . . 79
4.30 Captiva Island beach nourishment project and borrow area locations. 82 4.31 Captiva Island Model and Survey Losses . . 83
LIST OF TABLES
2.1 "Present" Wave Gage Stations and Locations, Spring 1989 6
3.1 Wave Height Information For Cape Canaveral, Indialantic, and Ft. Pierce.
Units of Hj/2 and H/ are (ft)s/2 and ft, respectively. These data
obtained by interpolation from adjacent CDN stations . 24
3.2 Wave Height Information For Jupiter Island, Delray Beach, and Hillsboro Beach. Units of Hj/2 and (H02)-4 are (ft)5/2 and ft, respectively.
These data obtained by interpolation from adjacent CDN stations 25
3.3 Wave Height Information For Pompano Beach and Hollywood/Hallandale.
Units of H;/2 and H/ are (ft)5/2 and ft, respectively. These data
obtained by interpolation from adjacent CDN stations . 25 4.1 Delray Beach Nourishment History . . 32
4.2 Jupiter Island Nourishment Summary. Multiple Reported Volume Per
Year Signifies Placement in Multiple Segments . 51
4.3 Hillsboro Beach: Percent Remaining of Original Volume . 61 4.4 Hillsboro Beach: Cumulative Volumes of Erosion Components 62
4.5 Hollywood/Hallandale: Profile Volume Changes With Time Between
Survey Events . . . . 71
4.6 Cumulative Component Percentage of Total Erosion . 76
4.7 Captiva Cumulative Volume Change . . 81
5.1 Sensitivity of k Parameter Compared to Dean Relationship Value, k, 85
Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science PERFORMANCE PREDICTION OF BEACH NOURISHMENT PROJECTS By
W. SAMUEL PHLEGAR III
Chairman: Dr. Robert G. Dean
Major Department: Coastal and Oceanographic Engineering
There is a definite need for improved interpretive and predictive capabilities for the performance of beach nourishment projects. Currently available analytical and numerical modeling procedures allow such predictions with one approach based on Pelnard Considere's theory which linearizes the governing equations. This theory, a combination of a continuity (sediment conservation) equation and a linearized dynamic (littoral transport) equation, results in a heat conduction analogy for the evolution of a shoreline planform which is out of equilibrium. This is a one-line approach with basic assumptions of profile equilibrium and parallel bottom contours; shoreline changes are thus represented by a single contour taken here as mean sea level.
Two linearized approaches are utilized to evaluate the performance of a beach nourishment project. An analytical approach presents the proportion of sand remaining in the region placed after any given time period and is based on an initially rectangular planform. The second method utilizes a finite difference solution which allows for much greater flexibility in representing actual project conditions at the various locations. The methodology is illustrated and evaluated by application to a total of ten projects around the State of Florida. The selection of projects was based on two factors including distribution around the State and most importantly, adequate monitoring information. Nourishment projects
included here and the respective dates of nourishment are Delray Beach (1973, 1978, 1984), Jupiter Island (1973, 1974, 1977, 1978, 1983, 1987), Indialantic Beach (1981), and Captiva Island (1981), Hollywood/Hallandale (1979), Pompano Beach (1983), Ft. Pierce Beach (1971), Hillsboro (1972), Treasure Island (1969), and Cape Canaveral (1975).
Each of these projects is analyzed with numerical modeling methods, and comparisons are made with actual measured volumetric changes. Analytical techniques are applied and compared on certain project locations. Results are encouraging, particularly for projects with accurate and ample monitoring information. The lack of follow-up information restricts the application as well as the interpretation on several projects. The need for improvement in monitoring procedure and archiving of obtained information is paramount for analysis of predictive capabilities. Sensitivity of the predictive performance is high with sediment information, ambient erosion rates, and wave characteristics. In light of the limitations in the methodology introduced, the results reflect an adequate procedure for prediction of shoreline evolution.
1.1 Purpose of Study
This thesis applies current capabilities to predict the performance of beach nourishment projects around the State of Florida. Significant work has been accomplished in analytical studies and numerical modeling in relation to beach nourishment projects. The focus here is on the adequacy of predictive capabilities in regard to nourishment and renourishment efforts based on application to actual projects. Much of the previous work has been based on Pelnard Considere's theory based on linearization of the governing equations which will be the general direction taken and foundation upon which the present study is based. Project performance data have been assembled to serve as a basis for evaluating the analytical and numerical modeling techniques.
Of the many nourishment projects over the past twenty years, few have been monitored to a satisfactory level for detailed analysis and evaluation of prediction methodology. However, the assessment of prediction methods requires adequately detailed comparison over the years after placement. Projects included in this study have at least a minimum level of performance information. The actual monitoring results are thus compared to the predicted performance to provide a measure of our prediction capabilities.
Nourishment is becoming the method of choice in coastal erosion control in Florida. The Florida Department of Natural Resources (FDNR) shows an accounting of $ 115.6 million spent toward nourishment during the period 1965-1984. From 1970 through 1984, there has been approximately 62 million cubic yards of sand placed on Florida's beaches. FDNR is currently asking the 1989 Legislature to appropriate $45 million for beach management projects for 1989-1990 fiscal year alone. Of this, $28 million is for new beach restoration and
another $10 million for renourishment. A demonstrated capability to predict adequately the performance of a potential beach nourishment project would therefore contribute significantly to the engineering, economic, and environmental assessment of the project. From the engineering aspect, the volumetric movement of sand from the project area and the associated time frames would be understood further. Coastal protection to properties and structures from the changing sand volumes both within and adjacent to the project area could be better evaluated. This would also enable more precise decisions concerning renourishment intervals, sand bypassing frequency, and maintenance placement. Sensitivity to sediment parameters might actually preclude the use of particular borrow sediments and reduce choices of sediment characteristics. Economically, overfills due to the planning now employed could be more accurately determined and thus significantly decrease the initial application costs. Proper advance planning would improve the economics involved with future renourishments as well as improve permitting/approvement delays. Finally, improvement in the understanding of the movement of the nourished material would allow greatly improved environmental assessment of the project area and adjacent areas.
This study is not intended to provide a highly advanced modeling capability in beach evolution, albeit a useful and effective one-line model will be discussed and demonstrated. The goal here is to analyze and assess our abilities to predict the performance of beach nourishment projects using existing technology. Modeling and applicable research in this dynamic area is exciting and due to the complex and variable "forces" caused by waves, a numerical approach is essential to realistic prediction.
1.2 Project Selection
Project sites to be studied were selected from many documented cases around Florida's coastline. Published information on many projects is severely limited to tabulated results which are inadequate for evaluation once again reinforcing the need for increased monitoring information. Thus, projects were chosen based on adequate construction information and monitoring results.
Adequate project information is centered around four specific information areas; construction, sediments, waves, and surveys. Construction placement density in planform is rarely in the idealized rectangular planform. It is important to represent the actual distribution of nourishment densities along the shoreline. In general, detail of distribution of sediments placed is sketchy yet significant to project performance. Beginning and completion dates, and construction delays, need to be reported and subsequently included in any analysis. Adequate information on the distribution, sorting, and size characteristics for both native and borrow sediment is necessary for increased understanding of results and performance. Wave forces are the key to shoreline change, and the analysis and representation of the wave climate at an area is very important. Adequate representation of the wave climate requires proper transformation to the breaking zone. Post project surveys provide the information to allow comparisons with predictions and need to be adequate in extent, planform density, and frequency following construction. Profiles should extend seaward to the limit of sediment motion for accurate volume determinations. The projects chosen vary greatly in all of these items, and each will be highlighted in its availability or absence.
Projects to be analyzed on the Florida east coast include Delray Beach, Palm Beach County; Jupiter Island, Martin County; Indialantic Beach, Brevard County; Hillsboro Beach, Pompano Beach, and Hollywood/Hallandale, Broward County; Cape Canaveral, Brevard County; and Ft. Pierce,St. Lucie County, and on the Florida west coast: Treasure Island, Pinellas County; and Captiva Island, Lee County.
2.1 Review of Methodologies
Both the analytical approach and numerical modeling serve as the theoretical background for prediction of planform evolution. There are no known solutions of any significance to the full equations of sediment transport and continuity. Pelnard Considere (1956) introduced the one-line theory utilized here in which the basic equations of sediment transport and continuity have been linearized and combined. This is the basic methodology followed in this study. This theory has been applied and researched extensively since its introduction and has proven an effective approximation to many shoreline applications.
The basic underlying assumption is one of profile equilibrium and associated constant beach profile shape. Complex conditions which exist at all locations are approximated and included in the solutions when possible; however, one-line theory requires some detailed limitations in application. Many analytical solutions have been developed for the combined equations, one of which will be highlighted subsequently. The limitations which exist in the analytical application limit its validity and acceptance in nourishment prediction. On several of the projects studied, however, the analytical results will be calculated and included in the performance comparison.
The numerical model basically represents the governing differential equations for shoreline response by their finite difference approximation. Parameters are adapted from each location to quantify nourishment attributes. The key here is to obtain realistic parameters for each specific location and include as many as possible in the numerical application. The procedure is basically as follows: grid network establishment over the computational domain, finite difference approximation of the governing differential equation, and solution
based on initial and boundary conditions. This equation is then solved repeatedly thereby advancing the solution forward in time. The "as constructed" project conditions serve as initial conditions for the computations which are then continued to times of available survey data used in the comparison.
The combined equations will be solved utilizing an explicit approach, as described in a subsequent section. The numerical model provides greater flexibility to more truly represent actual conditions. For example, known background erosion rates and repetitive and segmented nourishment events can be accounted for appropriately, as well as a more complete picture of construction and associated planform intricacies.
2.2 Relevant Parameters
Relevant parameters unique to particular locations include the depth of limiting motion, characteristics of native and borrow sediments, wavelength, local boundary conditions, local initial conditions, and background erosion information as determined by FDNR. Each of these parameters plays an important role in understanding the problem setting as well as in application of existing methodology. The approach to quantifying the relevant parameters is presented in the following paragraphs.
2.2.2 Wave Characteristics
A dominant term in the shoreline evolution is wave height, and any general choice of an average wave height without regard to season would be inadequate. Waves arriving from deep water are affected by many factors including refraction, shoaling, reflection, diffraction, friction effects and others. Topography, currents, and general specific bathymetry complicate matters as well and all together create a formidable system to model. Shoaling effects caused by wave propagation into shallow water are included here as well as a limited treatment of refraction.
The Coastal Data Network (CDN) wave data collected by the University of Florida are utilized (see Figure 2.1). Currently, eight stations around the coast of Florida table 2.2.2
Table 2.1: "Present" Wave Gage Stations and Locations, Spring 1989
STATION DEPTH(ft) LOCATION JACKSONVILLE 31.2 30 18' 00" N 81 22' 55" W
MARINELAND 32.8 29 40' 03" N 81 12' 17" W
CAPE CANAVERAL 26.2 28 24' 42" N (nearshore) 80 34' 36" W VERO BEACH 24.6 27 40' 30" N 80 21' 15" W
WEST PALM BEACH 29.5 26 42' 07" N 80 01' 42" W
MIAMI BEACH 21.3 25-46' 06" N 80 07' 23" W
CLEARWATER 16.4 27 58' 44" N 82 51' 00" W
VENICE 23.6 27 04' 26" N 82 27' 23" W
collect wave data at approximately six hour intervals. Included in this analysis are the relative depth of the gage, significant wave height, wave period, and percent wave energy distribution within various period bands. The initial wave information for the various project locations has been compiled from both CDN data reports and a 25-year nearshore wave evaluation published by CERC (Thompson, 1977). Currently, over three years of CDN data are retained on computer files at the University of Florida and are accessed for this evaluation. The wave data are represented by month, by location, and by spatial interpolation between data measurement stations. The significant wave height measurements are manipulated to obtain an effective breaking wave height at the shoreline, which serves as an important factor in the shoreline evolution. The seasonal effects are therefore included due to the monthly application. Wavelengths are summarily calculated at each of the project location from the dispersion equation utilizing the measured data. Nothing could substitute for actual wave data collected at the individual project locations; however, the CDN data allow actual data in the general vicinity to be interpolated to each site.
WATO St. Mary's Entrance
A -, S:---.., ,I'HA I TSA'I-- -: LUN AAMITON~A 0 1983 SEAT Y WAKULLAU AE C 8EYTa WAKUL TAYLOR g --T ALO1.1 R AS
Marineland' Steinhatchee DIX, C ALAC Ap LEVY
GULF AR VLUSI
CITUSI I LA
A HERNANDb--s PASCO Cape Kennedy
197C <2 IA 1977
I IA N A S i
IVenic ~ E --
CElearwI Aer ELAC Palm Beach
RA I; A
en \ Miami
Figure 2.1: Location of Coastal Data Network Stations maintained by the Coastal and Oceanographic Engineering Department at the University of Florida.
Simply stated, a general effect of wave action on a shoreline is the movement of sediment. Under circumstances particular to each location, the shoreline profile evolves toward an equilibrium shape. Under normal conditions, this shape is maintained through recovery processes in response to seasonal effects and storm activity not unusually severe. The shape is dependent primarily upon the wave climate and sediment particle size. Significant alterations in these factors force an attempt to reach a new equilibrium profile. When a nourishment event occurs, the beach is altered and thus set into motion to obtain equilibrium. This attempt is made in profile as well as the spreading out (diffusion) effects in the longshore direction. The effects of waves and currents drive the evolution toward the unaltered "straight" equilibrium shoreline. In the linear interpretation, the resulting planform is symmetric about the center point of the fill. This will be discussed in detail later. Essentially, the erosional problem after a nourishment event is due to a combination of these diffusion losses and natural underlying erosion present before the shoreline pertubation.
2.2.3 Background Erosion
An erosional problem is characterized by a background erosion rate; to a first approximation this erosion component will continue at the same rate superimposed on the "spreading out" losses due to the nourishment. Historical shoreline change can be caused by many factors and can also be accelerated or decreased over a period of years. Wave action is the ultimate cause of shoreline change; however, interaction with coastal structures, particularly jetties, provide the greatest effects on the Florida east coast. Inlet jetties and dredged navigational channels cause interruptions to the flow of sediment downdrift. Therefore, sediment that would have been transferred down the coast is removed from the sand sharing system. In areas with significant net directional littoral drift, this effect is of greatest significance. The downdrift shorelines then feel the effects of the volume deficit. Sea level changes over time will also cause shoreline alterations. A good approximation of the underlying rate of sand removal from a shoreline is a very important consideration, for without a background erosion, there would be no need for nourishment. The system con-
tinues to respond to the effects of this background rate after the nourishment application. Thus, the separation of background erosion effects from diffusion effects is a very important element in complete understanding of the shoreline's response to nourishment.
Historical shoreline positions have been digitized and compiled by FDNR for much of the coast of Florida. This information has been organized and combined from aerial photographs, maps and charts, and hydrographic and beach surveys, with information starting in the mid to late 1800's. Thus a long term background erosion rate or short term, should a change such as cutting an inlet or installing a coastal structure become involved, may be obtained. The data are referenced to FDNR monuments and further to State Plane Coordinates. Ground survey data are provided by FDNR, U.S. Corps of Engineers, U.S. Coastal and Geodetic Survey, state, county, city agencies, and others.
Seasonal shoreline fluctuations in Florida can be as great as 60 ft., complicating somewhat attempts to analyze the data and select an appropriate background erosion rate. Recognition of scatter and noise in the reported survey data within the project areas is important. Seasonal variations in the data may be compounded by unusual conditions associated with storm events. Other background rates will be referenced and applied to supplement the compiled data where considered appropriate. With the importance of an accurate background rate, sensitivity toward this factor is highlighted in many areas where the historical data are questionable or in contrast to other sources. Accuracy of the FDNR survey data is reasonable for the ground survey data with plus or minus ten feet the norm, while the aerial photos introduce a possible error of plus or minus 30 feet. In this light, the aerial survey data will not be included in determining the shoreline change. A long term rate is thus extracted from the data taking into account the possible sources of error and the limitations introduced by shoreline construction at inlets.
2.2.4 Local Conditions
Local conditions in the form of boundary and initial conditions provide the specifics of the model application. Each segment of coastline around the state of Florida is unique
in many ways. Several miles separating two areas can provide different sets of parameter application including differences in sediment characteristics, background erosion rates, and coastal structures. Numerical application is a powerful tool in this aspect in comparison to the analytical approach which is severely limited in its inclusion of many of these factors. Initial conditions on each project significantly affect the resulting performance with dominant factors including; length, volume, and the subtle effects of construction distribution in planform.
Boundary conditions provide projects with a certain uniqueness which must be addressed. The following discussion is based on figure 2.2 illustrating the various boundary conditions to be handled in the application section. Open coast beaches without structures, inlets or anomalies are the most adaptable to the basic methodology. The idealized condition for both numerical and analytical application is a straight shoreline with a rectangular beach fill extending into the ocean a distance Y, with longshore length 1. Also in this theoretical approach the shoreline is considered to be infinite in length with no interruptions. This is the idealized case upon which the analytical approximation is based. This is sketched in case 1 of figure 2.2. Erosion following nourishment in this case is equal to the sum of the background rate and the effects of planform evolution outside of the project limits. This case is well understood numerically by easy manipulation of the boundary conditions. The domain ends are far enough away from the affected area to not "feel" any effects and the shoreline displacements at the ends are set equal to zero. In reality however, a particular project may have features which differ substantially from those described above.
Shorelines requiring nourishment are in a state of erosion caused by many factors, both man made and natural. Also, constructed projects rarely approach the rectangular extension of beach. The current understanding of model application with regard to certain boundary conditions is limited. Utilizing a one-line model increases this lack of confidence in application. The introduction of littoral barriers such as groins, jetties at trained navigational inlets, or sinks such as the end of a barrier island all cause increased challenges
Case 2 Actual
BP Case 3
a P -0
a) QBP 0 b)0 C QBP 1
Figure 2.2: Three Possible Boundary Condition Applications: 1) Open coastline. 2) Structure on beach with minimal net longshore transport. 3) Inlet with a dominant transport direction
in interpretation and prediction. However, limited treatment in specific cases proves adequate. Specifically, figure 2.2 illustrates two cases requiring a different approach than the "ideal" case 1. There are two components of losses in this case, spreading out losses and interruption of the longshore sediment transport due to the presence of a jetty.
Case 2 is representative of a location without a dominant direction or significant volume of longshore transport. Transport through the jetty is assumed zero and the fill spreads in the direction away from the structure. At this end of the computational domain, the maximum grid cell is placed similarly to case 1, sufficient distance from the nourishment to assure zero transport into the cell. This results in an effective project length of twice that of the actual project. This approach will be utilized at Captiva Island and Treasure Island, both locations on the west coast. These areas are noted for the absence of a dominant transport direction and less wave intensity.
The third case illustrates the application to locations with natural or mechanical bypassing. The east coast of Florida is predominantly under the influence of a significant southerly net direction of sediment transport. The mechanics will be examined in chapter three, in the section on the numerical solution. The approach however is one of determining the bypassing quantity of sediment, QEP. A quantity of sediment is drained from the system and the shoreline approaches a normal angle to the incoming wave angle adjacent to the jetty. The transport deficit caused by the existence of the jetty forces this angle immediately downdrift of the downdrift jetty to be parallel to the incoming wave attack. The sediment erosion is the difference between the net longshore drift for the area minus the bypassed quantity. Two situations exist in this case. First, a bypass rate of zero indicates a deficit to the system of the entire ambient transport drift. The second option is the bypassing of a fraction of the nominal transport rate. As 22- approaches unity, the case is similar to case 2.
Initial conditions must also be prescribed, allowing representation of actual sediment placement in planform as well as sediment differences between the native sediment and the
borrow material. Sediment characteristics are a very important factor in any nourishment analysis or prediction. Median grain sizes and the sorting through distribution are included where the information is available. The information for both native sediment and borrow material is ultimately important in the resulting performance. Each project requires thorough analysis in this area, yet the records for most are severely lacking.
Finally, with the large scale of this analysis and the associated lack of highly detailed project information, it has proven instructive to look at parameter application in both a prognostic and diagnostic light. Thus, an attempt is made to first, learn from nature the attributes of the parameters and second, see if we can predict back to quantify an actual application.
The diagnostic approach will be centered around parameter sensitivity and the attempt to minimize deviation from the actual results. On the other hand, a prognostic approach is of value in the planning stage of any future project. Several parameters can be controlled and optimized while others are completely under nature's control. A more complete understanding of the variability due to an individual parameter will enable a more complete and confident plan for beach nourishment. Site specific local conditions in the form of initial and boundary conditions are therefore the basis for any approach.
3.1 Governing Equations
The two governing equations are the sediment transport and continuity equations. The background of each is essential to the understanding of the subsequent manipulation and application. As noted before, the theory was introduced in 1956 by Pelnard Considere. The methodology begins by introduction of the two basic equations of sediment conservation and sediment transport.
3.1.1 Continuity Equation
The continuity equation is a basic statement of sediment conservation where a gradient must be manifested by a change in the cross-sectional area of sand in the profile.
aQ a A
+ =A 0 (3.1) ax at
The change in sand area, AA, due to a profile displacement, Ay, active over a vertical dimension (h. + B) is depicted in figure 3.1. The assumption is therefore, in response to an accretion or erosion, a change in area is produced by a displaced profile and may be expressed as:
AA = Ay(h. + B) (3.2) This simple statement will serve as the first step in the method of calculation of nourishment volumes, both remaining and placed. This is followed by multiplication by the alongshore distance of interest. Thus the equation of sediment conservation results in:
= 0 (3.3) t (h.+B) 8x
Sediment is therefore conserved through the area of concern.
Depth of Closure
Figure 3.1: Profile Displacement In Response to Accretion or Erosion
3.1.2 Dynamic Equation
The second equation is the dynamic or sediment transport equation. Bagnold (1963) related available power in the surf zone to the total immersed weight transport rate. Introducing a relationship to the volumetric rate of transport, Q, and solving the relationship for Q, a sediment transport rate can be defined in terms of the energy flux in the longshore direction. It can be shown, by inducing breaking conditions of linear wave theory including utilizing the spilling breaker assumption, Hb = ichb (3.4) that the one dimensional transport equation is:
kH; sin2( ab)
Q = k(3.5) 16(1 p)(s -1)
where, in reference to figure 3.2,
* i = breaking wave criteria constant (0.78)
s= p./p,,: P, = mass density of sediment, p,, =mass density of water;s=2.65
p porosity (0.3 0.4)
g =gravity constant (32.2 ft/sec2)
Hb = breaking wave height at shoreline
*# = azimuth of outward normal to shoreline
p = azimuth of the shoreline's general orientation defined by the baseline
e = azimuth from which breaking wave originates
o k = transport coefficient (depends on sediment size)
3.1.3 Combined Equation
The combination of the continuity and sediment transport equations provide the framework for further solution. First, equation 3.5 is differentiated and the result linearized along with the equation for f. Assumptions for linearization include: uniform wave height distribution along the shore, small gradients of the shore planform, and a small wave approach angle ( ).
r ay a 8y
8 = IL- tan- ) ( (3.6)
2 ax 2 ax which, when combined with equation 3.3, becomes the partial differential equation governing the planform evolution: ay _2
G a2 (3.7) t = 22
This equation is seen as the typical form of the "heat conduction" equation, and is the combined linearized equation for which many analytical solutions exist (Larson 1987). The term G is described as an alongshore diffusivity term and depends strongly on the breaking wave height,
G k;V.(3.8) =89(s 1) (1 p) (h. + B)
--Shoreline Reference Base Line
Figure 3.2: Sketch showing shoreline orientation and parameter explanation
This single term includes several significant parameters including wave height, sediment characteristics, and closure depth. Resulting evolution is strongly associated with this term justifying the "alongshore diffusivity" terminology.
3.2 Analytic Solution
One of the most valuable solutions to the diffusion equation in coastal engineering is that for an initially rectangular nourishment project. The resulting analytic solution for this case can be expressed with y(x,t) being the displacement offshore and a function of alongshore distance, x, and time, t:
Yr i 12x\1 r ,2x y(x, t) = jerf I '(2x+1)] -erf ( 1) (3.9)
2 4zU -,fG 4 ,fd- I Here "erf"denotes the error function which can be found tabulated in standard mathematical references, (e.g. CRC, 1984), and is defined as: erf (z) = 2 fexp(-u2)du (3.10) The term 1 is the length of the project in the shoreline direction, and Yo is the initial shoreline displacement resulting from the nourishment extension seaward. The evolution of such an initial case is shown in Figure 3.3. The following parameter establishes similitude between two projects.
Thus two projects with the same parameter shown above, will exhibit similarity in evolution and the resulting percentage volume changes.
Based on equation 3.7 it is possible to develop an expression representing the proportion of sand remaining in the location placed, M(t), as:
M(t) = 2 V/- t (exp (/2x/Oi)2 1) + erf (1/2v/di)) (3.12) /p1) (/2 ) 3.2
From Figure 3.4, proportions of the fill remaining in front of the original location placed can be established based on the dominant similitude term, equation 3.11. Similarly, a solution
DISTANCE FROM ORIGINAL
SHORELINE, y (ft)
- / I7
Planform After 3 Months
ALONGSHORE DISTANCE, X (miles)
Figure 3.3: Shoreline evolution for initially rectangular fill, I = 4miles, H 2ft., Width = 100ft. (from Dean, 1988)
0 1 1 1 1 1 1 1 I = Time After Placement G= Alongshore Diffusivity ninll Asymptote Planform 5- M=1
Figure 3.4: Proportion, M(t), of fill remaining within original project limits.
-Jzw Z 0
0. 2 .0.
< 0 CL. 2 -J0.
can be developed for the situation of a project located downdrift of a partial or complete littoral barrier. Several examples of this application exist on Florida's east coast.
1iVG- / 1 f-) 1) (1 F)Q,,t
M2(t) = I exp -- + erf (-//ft) F (3.13) The percentage of the longshore transport, Q,, bypassed is F with (0 < F < 1). The original volume placed is represented by Vo.
3.3 Numerical Solution
Solutions developed numerically allow much greater flexibility in comparison to the above due to the ability to include numerous process parameters. The numerical solution is structured with a finite difference representation of equation 3.7. This results in the following explicit form:
y+1 =y' + T2 (nf-1 2y? + yi4+) (3.14) Where the superscripts denote time level (n+1 is the next time step) and the subscripts represent the grid in the network illustrated in Figure 3.5.
Boundary conditions representing the individual projects are specified and incorporated into the solution. Introduced in Chapter 2 and illustrated in figure 2.2, the explicit method of solution utilized in this study allows basically two types of boundary conditions. The first is where the value of the shoreline at the project ends is specified and either held fixed or allowed to vary with time in a prescribed manner. As an example, ymaz and yo on an open coast project without protruberances would be set equal to zero and selected at sufficiently large distances from the fill that the evolution would not be affected by the zero flow condition.
Many projects in Florida are, however, affected by the presence of a jetty. When this condition exists, the flow of sediment is restricted it that particular location and must be treated accordingly. This second type utilizes the procedure of imposing a sediment transport rate, Q, at the boundaries. Examples here would be a zero transport due to a total littoral barrier and a low or "leaky" jetty which allows a certain amount of sediment
Reference Baseline for Shoreline Ax Ax
Hy -1 f
~----- Y-----~ I +1
Figure 3.5: Schematic representation of shoreline for numerical application.
A ~ ~ i+1 --------
movement through the structure. For purposes of further exploring this boundary condition, one of the assumptions listed earlier of assuming negligible wave approach angles must be modified. Also, the approach angle is retained in a sine term and not a cosine term which would approach unity with small angles. Equation 3.5 and equation 3.8 can be reduced to (Q QiP) = G(h. + B) 2 (3.15) The left side of equation 3.15 allows the alteration of the sediment deficit due to the amount bypassed, QBP. Should the approach angle be zero, the equation would reduce to Q equaling QEp and the approach would be to follow case 2 of figure 2.2. The method followed in modeling of this case would be to double the computational domain and simply sum over one half its length. This forces the no transport condition at the jetty, the midway point. Other physical conditions that could require representation include: groin location, natural inlets, and ends of a barrier island. Each of these cases will be discussed in detail as the need arises for their application.
The defense posture of the beach orientation alluded to earlier is simply due to the shoreline's attempt to reach equilibrium and meet the condition of zero flow through the jetty. To achieve zero transport, Q0 = 0, the sine term must vanish, therefore ac, equals 0 and the shoreline orientation is parallel to the wave crests or at a ninety degree angle with the wave rays.
A limitation is noted here due to the neglect of two very important terms, wave diffraction around the coastal structure and seasonal reversals in dominant wave direction and associated longshore transport direction. It would be optimal to include procedures for the diffraction effect as well as the prediction of planform changes due to reversals.
3.4 Wave Parameters
A description of the various wave parameters and associated determination is necessary. The CDN wave gage network collects data from relatively shallow water submerged gages. To obtain a wave climate representative of that at the coastline, the following procedure is used.
The equation for conservation of wave energy flux (Dean and Dalrymple, 1985) with the subscript b denoting breaking depth conditions is [EC, cos(6 a)] = [ECg cos(P a)i (3.16) with the energy term defined by
E = -pgH2 (3.17)
The flux expression is equated at the breaking depth and at the depth of the CDN station. Again assuming unity for the cosine terms and applying the shallow water approximation for wave celerity, shown subsequently, as well as the spilling breaker assumption, equation 3.4, the resulting expression is equation 3.19.
C = /g-h (3.18)
H/2- H2Cg (3.19) where
C = [ + 2kh (3.20) 2g= snh 2kh
Here, C, is the group velocity at the offshore reference location. Longuet-Higgons (1952) showed that waves in nature are represented well by the Rayleigh Probability Distribution. Equation 3.19 can be rewritten as H12 = F(T.) H2 (3.21) where
F(T.) C9 (3.22) By applying a Rayleigh wave height distribution relationship, the following sequence is followed for solution of the resulting breaking wave height; H = p(H)F(T)H 2dH (3.23) = F(T,) H2p(H)dH (3.24)
Table 3.1: Wave Height Information For Cape Canaveral, Indialantic, and Ft. Pierce. Units
14 2 dare (ft)2 and ft, respectively. These data obtained by interpolation from adjacent CDN stations
MONTH CANAVERAL INDIALANTIC FT. PIERCE 5/|2 H |2 0.4 H (P 2 H02o H |2 H |y.4 JAN 5.97 2.04 8.95 2.40 12.51 2.75 FEB 4.96 1.90 7.85 2.28 9.77 2.49 MAR 10.59 2.57 14.62 2.92 16.80 3.09 APR 3.66 1.68 4.84 1.88 5.91 2.04 MAY 4.32 1.80 6.14 2.07 8.61 2.37 JUN 2.52 1.45 2.96 1.54 3.37 1.63 JUL 1.44 1.16 1.88 1.29 2.13 1.35 AUG 2.73 1.49 2.75 1.50 2.32 1.40 SEP 4.48 1.82 3.96 1.73 2.78 1.51 OCT 10.45 2.56 11.63 2.67 15.19 2.97 NOV 8.53 2.36 11.56 2.66 13.29 2.81 DEC 9.15 2.42 11.18 2.63 12.71 2.76
We know through the distribution and resulting relationships with H,.,,, and H, that
H, = 1.416 [H]1/2 (3.26)
Substituting for H2 and averaging over the data collection period t,, results in
n= 0.496 = F(Tn)H2 (3.27) n=T1 Atn
and H8m denotes the significant wave height over the nth time increment, Atn The data
are read from files and analyzed based on the time between records. Therefore, for each
record taken (usually every 6 hours), F(Tn) is calculated and inputted along with Hn into
Tables 3.4, 3.4, and 3.4 present the developed effective wave heights from the above
analysis at each of the east coast sites. The data are listed by month and from north to
south along the eastern coast.
Table 3.2: Wave Height Information For Jupiter Island, Delray Beach, and Hillsboro Beach. Units of H; and (H;/2)-4 are (ft)5!2 and ft, respectively. These data obtained by interpolation from adjacent CDN stations
MONTH JUPITER ISLAND DELRAY BEACH HILLSBORO BEACH
H_(_)_-__H (b )0-4 H2 H ; 4
______ bHt b bH2). b~
JAN 12.30 2.73 10.02 2.51 8.47 2.35 FEB 6.69 2.14 3.67 1.68 3.38 1.63 MAR 11.58 2.66 6.68 2.14 6.34 2.09 APR 5.19 1.93 3.72 1.69 3.07 1.57 MAY 9.06 2.41 7.55 2.25 6.15 2.07 JUN 3.15 1.58 2.54 1.45 2.23 1.38 JUL 1.57 1.20 0.97 0.99 0.87 0.95 AUG 1.47 1.17 0.72 0.88 0.71 0.87 SEP 1.77 1.26 0.83 0.93 0.78 0.91 OCT 19.16 3.26 17.22 3.12 13.26 2.81 NOV 9.52 2.46 5.50 1.98 4.91 1.89 DEC 10.89 2.60 7.97 2.29 6.94 2.17
Table 3.3: Wave Height Information For Pompano Beach and Hollywood/Hallandale. Units ofH and (Hi/)-4 are (ft)5/2 and ft, respectively. These data obtained by interpolation from adjacent CDN stations
MONTH POMPANO BEACH HOLLYWOOD/HALLANDALE
______H;' (H; 0) 0.4 H b 2 (H;/2) 0.4
JAN 8.02 2.30 5.88 2.03 FEB 3.29 1.61 2.89 1.53 MAR 6.25 2.08 5.78 2.02 APR 2.89 1.53 2.00 1.32 MAY 5.75 2.01 3.80 1.71 JUN 2.14 1.36 1.71 1.24 JUL 0.85 0.94 0.71 0.87 AUG 0.70 0.87 0.69 0.86 SEP 0.76 0.90 0.68 0.86 OCT 12.13 2.71 6.67 2.13 NOV 4.74 1.86 3.92 1.73 DEC 6.64 2.13 5.22 1.94
3.5 Remaining Parameter Summary
3.5.1 Wave Refraction
It can be shown that direct application of the linearized equation 3.7 overestimates the transport losses due to the neglect of wave refraction. Specifically, the factor G in equation 3.8 should be modified by multiplication by a term R presented below. In estimating an effect due to refraction we implement Snell's Law to determine a ratio between the wave celerity at the location of the depth of closure and at the effective breaking depth.
sin(8 ab) = Cb/C.sin(f a.) (3.28) R = Cb/C. (3.29) Here, C. is the wave celerity at the depth of profile closure and Cb that at breaking. This modification to the diffusivity factor will be applied to both analytical and numerical applications.
Closure depth, defined as the limiting depth beyond which no sediment transport occurs, is difficult to establish precisely. Representative values of 18 and 10 ft. for the Atlantic coast and Gulf coast respectively are assumed initially. These will be altered where information at a particular project suggests otherwise. Volume calculations depend highly on this term since we assumed a uniform profile movement.
3.5.2 Sediment Size
Dry beach width gained per unit volume of nourishment material is directly related to borrow sediment size. A coarser fill sediment will result in greater beach width than smaller sediment sizes, and more of a finer nourishment sand will be required to fill out the offshore milder associated slopes. Additionally, greater amounts of sediment transport will occur for finer than coarser sediment. Relationships exist which are developed by analytic models suggesting relationships between k and sediment properties. The Army Corps of Engineers presents a relationship of k with the sediment fall velocity parameter. A relationship between sediment size and k values has been suggested by field studies (Dean,
0 0.5 1.0 DIAMETER, D (mm)
Figure 3.6: Plot of sediment diameter and the k value (after Dean, 1988)
1988), and is utilized in estimating a first approximation to the k value. This relationship is shown in figure 3.6. Where inadequate sediment information exists, the constant, k =0.77, will be applied (Komar and Inman, 1970). These guidelines are followed for initial constant application and both k and the closure depth will be studied by sensitivity analysis in several situations where indicated by data.
3.6 Model Summary
The model is a basic one-line numerical iteration based on input of site specific parameters. The computational domain is chosen based on the project initial conditions and boundary conditions. The resulting planform following construction is overlaid onto the domain and allowed to evolve based on the methodology above. It can be seen that the significant parameter in the numerical application is the following term seen in equation 3.14:
Ir = (3.30) AX2
In this simple term, we have most of the major "movers" of sediment following a nourishment event. It can be shown that in order for numerical stability, I' <0.5.
The model is set up to churn through the months based on the changing wave conditions as well as days in the month. The time increment At is set at 86,400 seconds or one day. Thus, the model acts on the material placed on the previously "straight" shoreline in intervals of one day. At times of interest, the model outputs volumes remaining within original project limits.
The grid is laid out according to figure 3.5 and set up with applicable grid widths ranging from 400 to 1000 ft. in the longshore direction. Wave heights are transferred to an applicable breaking height appropriate for our analysis (equation 3.27). The information calculated at each CDN location is interpolated to the project location of interest. Background nourishment information is input at known locations in reference to monument location and interpolated to the grid location for the model. These values are then allowed to act on the remaining volume and output is provided at the various times of interest. Multiple nourishments are handled by adding the renourishment amounts from the new construction to the existing "remains" from the action on the previous nourishments. It is therefore a cumulative process of determining the final resulting planform. Volumes are calculated by multiplication of the sums of the extended beach lengths (seaward) times the grid width, and finally by the depth of closure. Again, the assumption of a uniform movement in profile is utilized.
The total erosion from the nourishment fill area will be the sum of the background erosion rate and the spreading out losses due to planform evolution. In the same analogy, the total littoral transport rate is due to the addition of the localized littoral transport and the rate due to the nourishment. The principle of superposition holds due to the linearity of the model.
Finally, sensitivity information is collected by allowing the important parameters to vary over the life of the project to obtain their influence and effect on the projected volumes. Results from the model volume calculations and sensitivity analysis will be reported for each project.
PROJECT APPLICATION AND RESULTS
Figure 4.1 shows the location of each of the ten projects evaluated here. Each project will be summarized below including specific nourishment characteristics, local conditions, historical shoreline changes, and parameter evaluation. Finally the actual versus predicted volumetric changes will be presented as well as appropriate sensitivity analyses.
4.2 Application and Results for Each Location
4.2.1 Delray Beach
Delray Beach is located in Palm Beach County on the southeast coast of Florida approximately 50 miles north of Miami (Figure 4.2). Five miles to the north is South Lake Worth Inlet while Boca Raton Inlet is 6.7 miles south. The city began an erosion control program in 1973. A reduced littoral drift environment exists at Delray Beach due to South Lake Worth Inlet in combination with shoreline defense structures just north (updrift). Vertical seawalls, concrete block revetments, and coral rip rap were installed in the 1960's to prevent further damage and erosion. Erosional pressure was thus simply being forced further south. Background erosion rates from DNR data range from an erosion of 5.8 ft/yr to an accretion of 1.4 ft/yr. These data are average changes over a 25 year period prior to nourishment 1945-1970, and are shown graphically in figure 4.3. The location of the DNR monuments are located in the grid network for the project and the associated shoreline change rate inserted with a running average of 5 points used to determine each grid point rate. The average rate of the period is approximately 1.5 ft/yr of erosion in the project limit domain. Variation of the background rate for approximately 4 1/2 miles north and
WALeONE -. -ASSAU
4 L--LON HAMILTON4
A 'A --, A S 1 1
I BERTY WAKULLA BA E ~AYLORAe
GULF FRA Al
OI CFO D \ST. I CHRIST JOHIIS
GULF R VLUSI
--- ODRA1GE 6Cape PASCO ---R LOSSCEOLA
C- RIVE Treasure A.AKE r's- Ft. Pierce LEMANATEE HARDEE NOR C SLCa
-- 6 LUCIE H AE SOTHOa
OKEECHO--- sa CHAR TTC AEj
Captiva L E NDRY rMtt I$LACW Beach Island- i Hillsboro
--- E Hallandale ONR~t DD
Figure 4.1: Location of beach nourishment projects to be studied.
south of the construction limits is also seen in the figure. A severe littoral environment thus existed and a long range solution was needed to replace the "quick fix" solutions. A long term beach nourishment solution was selected and an adequate sediment supply for borrow material was located directly offshore.
The erosion control program at Delray Beach thus began in 1973 with 1,634,513 cubic yards (CY) of borrow material placed on a project length of 13,850 feet. Following the rule of thumb in which 1 square foot of change in beach area equals 1 CY of sediment, the shoreline should be extended an average of 118 feet. Placement of sediment was not attempted as a rectangular (in planform) shape, and the actual volume addition distribution at each grid coordinate was replicated in the model (Arthur V. Strock & Associates (AVS), 1975). Initial design called for adequate profile monitoring and renourishment intervals with the original nourishment completed in August of 1973.
The first renourishment was undertaken from February through July of 1978 to increase storm protection and recreational benefits to those of the original nourishment. Approximately 70% of the original nourishment was found by survey to remain in the project area. A quantity of 701,266 CY of sand was placed in two site specific areas: One near the southern end of the city and one near the north, both within the original project limits (see Figure 4.2). This fill was estimated as essentially rectangular for modeling purposes (per conversation with Coastal Planning and Engineering, Inc.).
A second renourishment was constructed in September and October 1984 within the original nourishment construction limits. A volume of 1,311,006 CY of offshore material was pumped to the approximate 14,000 ft project length. A major storm on Thanksgiving Day, one month after completion, caused major damage along Florida's east coast. Effects on the upland beach and facilities at Delray were minimal due to this continued erosion control program. From the outset of the initial nourishment, 14 profile surveys have been performed and reported. This project monitoring effort far exceeds the normal program attempts. Detailed reports for the various monitoring studies are provided by Coastal
Table 4.1: Delray Beach Nourishment History DATE QUANTITY (CY) LENGTH (ft.)
8/73 1,634,513 13,850
7/78 701,266 6,746 2,225
8/84 1,311,006 13,850
Planning and Engineering (CPE). Table 4.2.1 is a summary of the Delray nourishment history.
Wave height information is interpolated from CDN data stations off West Palm Beach and Miami. The resulting effective wave height is 1.99 feet with an average period of 6.4 seconds. The model boundaries are maintained at sufficient distances from the projects so that the fixed shoreline boundary conditions do not affect predictions, following the case 1 description presented in chapter two (figure 2.2).
Figure 4.4 is a comparative plot showing model and survey volumes over a 14 year period from 1973 until 1987. This format will be utilized in most of the project result presentations, with cubic yards remaining on the y-axis versus increasing time from original sand placement of the x-axis. Each of the subsequent renourishments are easily seen in the figure by both model predictions and actual survey data. The closure depth has been reported as 15 feet for the Delray Beach area which relates to a value of 24 feet for (h. + B). Borrow material was much finer and slightly poorer in sorting than the native sediment with a reported size in the range of D50 = .23mm. The k factor used is therefore 1.3. This combination of parameter input resulted in a standard deviation of 242,000 CY, a value quite favorable for a 14 year period with multiple nourishments.
Sensitivity of model prediction performance in regard to the two parameters k and (h. + B), is plotted in figure 4.5. This is a contour plot showing, for ranges of k and (h. + B) values, the resulting variance. A minimum standard deviation of 90,000 CY and thus the "best fit" over the study period resulted from a k factor of 1.3 and a (h. + B) value
of 27 feet. With a prescribed (h. + B) value of 24 ft., the minimum variance occurred at a k value of 0.6. The model results data point shows the comparative variance due to model values of (h, + B) = 24 and k = 1.3. This data point and the minimum variance point are marked in the figure. The plot clearly shows a "valley" of variance values, beyond which the variance increases. To explore this sensitivity measure one step further, the variance is plotted versus the k parameter alone, figure 4.6, keeping (h. + B) constant at 27. The minimum variance results from k = 0.6 and the figure illustrates the resulting increase in prediction deviation away form this k value. It is interesting however, that with an increase to 27 for (h. + B) (which is considered an average value for the east coast of Florida) the minimum k is precisely that predicted by the relationship in figure 3.6. This diagnostic approach could be very important in future nourishments at this area.
Finally, ignoring refraction effects results in a much over- estimated volume depletion for the 14 year period. This is seen in figure 4.7 and is the basis for introducing a refraction application into the modeling scheme.
4.2.2 Cape Canaveral
The Cape Canaveral Project area is on the Atlantic Coast of Florida just south of the southern jetty at Port Canaveral (figure 4.8). This area was in a state of accretion until construction of the inlet in 1951 and associated training jetties from 1951-1954. Studies of the littoral environment suggest the existence of the jetties create a near complete littoral barrier for any southward sand movement (Hunt, 1980). The area south of the inlet eroded at an average rate of 5 feet annually from 1954 to 1965 (USAE, 1967). A nourishment project was begun in June of 1974 and completed in March, 1975 to alleviate the erosional stress on the area. Also indicated in figure 4.8 are approximate shoreline changes through seven years taken from aerial photographs (from Stauble, 1984)
The nourishment sediment was taken from the Trident Submarine turning basin, sampled adequately at appropriate intervals, and analyzed by standard sieving. Pre-construction and borrow material phi sizes are 2.00 and 1.59 respectively corresponding to .25 mm and
Figure 4.2: Delray Beach area and project locations
y LimitR-1 75
North Limit r1984 + 1973 Construction
1978 North Construction R-180
-South Limit R-189 1984 + 1973 Construction
DELRAY BEACH AREA BACKGROUND EROSION
GRI CCAINT ac:
-1 S t6 1 L 260 A 363 A 41 A 50 55 60 6L 710 A 8 L 9 1 D
GRID COORDINATE BACK EROS
Figure 4.3: Delray Beach area background erosion: shoreline change plotted at DNR monuments for 1945 -1970
DELRAT EFACH PROJECT: MODEL & SURVEY
MODEL .... .....SURVEY
Project completIon date Initial design volume (CY
8/1973 1,634,613 7/1978 701,266 8/1984 1,311,006
2800000 2520000 2240000 1960000 1680000 1400000 1120000
Effective wave eight 1.99 ft.] Sediment size 0.23 mm (h. + B) : 24 ft.
12 24 316 46 A 712 91 6 108 120 132
MONTHS AFTER INITIAL PLACEMENT
144 16 168
Figure 4.4: Delray Beach: model prediction and survey volumes
32. 31. 30. 29. 28. 27. 26.
25. 2q. 23.
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7
Figure 4.5: Delray Beach: parameter variance plot; contours of variance calculations X1010.
DELRAY BERCH: K FACTOR SENSITIVITY
I ro dct completion Initial deslg volume (CY)
r /I17 I 0 I6
fective wave height :19 t SdImeut se: 0.28 mm (h. + B) : 14 ft.
24.0 D 21.0
LI-I z 15.0
9.0 6.0 3.0 0.0
0.2 0.1 0.6 0.8 1.0 1.2 1.2 1.6 1.8 2.0 2.2 2.14 2.6 2.8
Figure 4.6: Delray Beach: k parameter variance with constant (h. + B) =24.0
DELRAY BERCH: REFRACTION EFFECT 200000
2520000 --- -- -----.MODEL W/O REFR
Project completion date Initial design volume (CY
8/1973 1,634,513 7/1978 701,266 8/1984 1,311,006
Efctive wave hil T-4k'- Sediment size 0.23 mm (h. + B): 24 ft.
1973 1978 1984
0 1 2 36 t8 60 2 9 108 120 132 4 --6 168 MONTHS AFTER INITIAL PLACEMENT
Figure 4.7: Delray Beach: refraction effects on prediction of evolution
.33 mm. The borrow sediment size relates to a k factor of 1.1. The project length extended 2.1 miles south of the inlet. Approximately 2,300,000 CY of sediment was pumped onto the project area with a majority located near the jetty end, the northern limit of the computational grid. Average wave heights of 2.01 feet affect the coastline with effects due to the large shoal of the Cape not taken into effect. The approximate size of the shoal is 100 million CY and is certainly capable of introducing refraction and diffraction effects. The CDN station offshore of the Cape provided the wave characteristics.
The boundary condition applied here is categorized by the case 3 example in chapter 2, figure 2.2, in response to the inlet and jetties. An approximation to the longshore transport rate is necessary for this application. Estimates of net longshore sediment transport rates range from 250,000 CY/yr (Walton, 1976) to 360,000 CY/yr (Army Corps of Engineers) both in a southward net direction. A subsequent forced slope immediately south of the jetty follows from the boundary condition illustrated earlier. An evaluation of erosion data (Dean and O'Brien, 1987) can be interpreted to show a volumetric rate of erosion south of the inlet of approximately 200,000 CY/yr. It will therefore be assumed here that (Q QBP) = 200,000 CY/yr. This results in an average wave angle of 5.8 degrees also indicating the resulting defensive posture angle of the shoreline south of the jetty. Calculations based on a drift rate of 350,000 CY/yr result in an approach angle of 9.3 degrees.
The responsiveness of the model to these two approach angles is illustrated in figure 4.9. The difference in prediction of volumes remaining at 12 months is 104,749 CY and at 72 months is 622,724 CY. An accurate littoral transport rate is obviously a key parameter in project application near structures.
Project profiles were taken before and after construction at project specific profiling locations tied into DNR monuments. Results of four monitoring efforts were presented over the course of the first six years after project completion. These are seen in figure 4.9 and can easily be compared with the prediction results of model application. The first monitoring volume at two months following placement only included the volume change above mean
water. It therefore reflects the initial sorting and movement offshore of finer material. The offshore volume change must be included for comparison in this analysis. The second monitoring report at eleven months reported a volume loss of 269,000 CY. Variance from the "5.8 degree" modeling prediction is 94,000 CY or approximately four percent of the original placement. Aerial photography and survey results at 66 months following completion show 68 and 60 percent respectively of the material placed remaining as shown on the figure. The survey results reflect a much greater volume remaining within the project limits. A difference of over 600,000 CY between prediction, "5.8 degree", and survey volumes results in the 66 month period following placement. It is possible, yet not incorporated, in the numerical application that an over-prediction would result if the nourishment volume does not replicate the original shoreline. Should the seaward extension be less, the fill would assume a more advanced state of evolution and thus less erosion compared to a fill out to the pre-inlet shoreline. Improvement in the modeling of this possibility is needed.
The erosion during the six year monitoring study was concentrated in the center half of the project boundaries which could be due to the sheltering effect of the Cape's offshore shoals. This can not be represented in the modeling effort, but introduces possibilities for future research. Sources of survey error were introduced in the 12 month survey report from -10 to -20 ft. depth. Volumes reported are therefore only calculated out to a depth of 10 ft. The model assumption is a profile shift out to the depth of closure, assumed here to be -18 feet. Should a positive volume exist in the error range of depths, the 12 month deviation could conceivably diminish.
Analytical prediction is also shown in figure 4.9. Close agreement with the 5.8 degree case is due to the introduction of an F factor of .33 and an ambient transport of 300,000 CY in to the analytical expression. Analytical results should reflect the finite difference predictions in cases where there are no additional nourishments. The variances shown in the figure are inherent due to the initial non-rectangular nourishment scheme with placement of larger volumes near the jetty.
High Water Lines
---- 2/14/43 Historical -- 7/17/73 Pre-Nourishment m
---1/9/75 Post-Nourishment.: >
---9/30/80 7 Years After
Figure 4.8: Cape Canaveral: project and borrow area locations. (after Stauble, 1985)
CAPE CANAVERAL VOLUME PREDICTION
2500000 -- -MODEL 5.8 DEG
MODEL 7.9 DECG Project completion date: 3/1975 225000(, .N. .. ICAL VOLDJ Initial design volume: 2,300,000 CY Project length: 11,088 ft. Effective wave height : 2.01 ft. Sediment size: 0.33 mm
(h. + B) : 27 ft.
NQ. Qnr: 200,000 CY
(_) 350,000 CY
I 750001 Approach angle : 5.8 deg. z 19.8 deg.
Cc I ODOGOU,
0 ---- --r -- r--- F --0 12 24 36 48 60 72 84 MON[HS AFTER INITIAL PLACEMENT
Figure 4.9: Cape Canaveral: survey volumes and resulting model volumetric changes for
two cases of varying transport rate.
4.2.3 Indialantic Beach
Indialantic Beach is located on the Atlantic Coast in Brevard County approximately 30 miles south of Cape Canaveral (Figure 4.10). The nourishment project started in October of 1980 and was completed by January, 1981. A total of 255,139 CY of sediment was placed on a project length of 2.17 miles. The material was obtained from the Port Canaveral submarine turning basin for use during the Cape Canaveral beach nourishment project in 1974, and had a similar mean size but a poor sorting in relation to the natural sand. A reported sediment size of D50 = .4 mm translates to a k factor of 0.9. The planform deviated significantly from a rectangle and original input information varied over the five fill "zones" (5 to 43 ft of beach width increase, see figure 4.11). This figure typifies the model setup used throughout the thesis in regard to grid layout and beach extension. Indialantic construction volume is very low in comparison to the other projects in both dry beach extension and volume/ft of project length (Stauble, 1986).
Three project profile lines and two control profiles outside of construction limits were established for the monitoring effort. Post construction profiles were taken twenty-four hours after placement of fill segments, whereas most nourishment projects are surveyed following the entire fill placement. This difference allows calculation of the initial sorting losses per fill zone. Also, the surveys extended seaward only to depths of 5 to 10 feet. These two effects introduce problems in the analysis of the resulting performance. The remaining volume of sediment located in the zone from the end of the profiles to the closure depth is not known and therefore cannot be taken into account.
Boundary conditions were similar to Delray Beach in that the computational area was set large enough so that there would be no influence within the construction limits. Results from Cape Canaveral and Vero Beach CDN included a representative wave height of 2.22 ft. and a period of 8.11 seconds. Figure 4.12 includes plots of the predicted model volumes and analytical volume results against the survey data at the various time periods. The large variance from the predicted results are assumed to be created by the unaccounted volume
beyond the profile survey depths and a poorly known background erosion rate. Erosion rate is set at zero for the project length in figure 4.12.
Stauble (1986) suggests a shoreline retreat rate of 5 ft/yr before 1960, and a relatively stable shoreline thereafter, while (DNR) rates for Brevard County vary significantly over this area of the coast. The background erosion effect is heightened on low construction volume projects in regard to the total loss calculations. A look at the sensitivity of this parameter is therefore warranted. To understand the importance of the background erosion rate with this small project volume, figure 4.13 plots model results produced with varying rates (+2, 0, -2, and -5 ft/yr). A significant variance results with a relatively small change of 2 ft/yr, while a change of 7 ft./yr. (+2 to -5 ft) results in a difference at 6 years of 500,000 CY (roughly twice the initial project volume). It also brings the prediction closer to the survey data; however they should never theoretically meet due to the neglect of sediment in depths greater than the survey extent.
4.2.4 Jupiter Island
Jupiter Island is located on the Atlantic Coast in Martin County approximately 80 miles north of Miami (Figure 4.14). The Town of Jupiter Island is 6 miles to the south of St. Lucie Inlet and has experienced significant erosion over the past one hundred years. Aubrey (1988) notes that the narrowing of the continental shelf in this region coupled with the lack of sheltering from the Bahama banks create a higher erosion potential than areas to the north (wider shelf) and south (sheltering). This is also reflected from the tabulated effective wave heights in chapter 2. Beginning in the mid-1950's, attempts were begun to stabilize the beach and have included some 8 million CY of sediment along with miles of seawalls, revetments, and groins. The first attempt at beach nourishment was completed in 1957. Material was pumped to the beach from Hobe Sound intermittently from 1957 through 1963. The sediment quality was suspect as bay sand is normally of poorer quality, smaller size, and contains higher organic and fine contents compared to other sources. Median grain size of the natural beach sand prior to 1963 was .29 mm.
BORROW AREA 0
R-132%9 PROJECT SITE Melbourne
Figure 4.10: Indialantic Beach: project location and borrow site. (after Stauble, 1985)
91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110
Ax=-600' ALONGSHORE DIRECTION, X (Computation Coll Numbers)
Figure 4.11: Indialantic Beach: bar diagram of planform and grid cell network after fill
INDIALPNTIC PROJECT: MODEL & RNRLYTICRL
4 36 48 MONTHS AFTER PROJECT
Figure 4.12: Indialantic Beach: model and analytical prediction showing survey volumes.
Project completion date: 1/1981 Initial design volume: 255,139 CY Efective wave height : 2.22 ft. Sediment lsei: 0.4 mm
-- -- -- -- -- -- -
h. + B) :
........... -...... ..... -.............. ....-
INDIILANFIC PROJECT: EROSION HR TES
+2 FT Project completion date: 1/1981
0 FT Initial design volume 255,139 CY
- 2 FT IEffoctive wave height : 2.22 ft.
Sediment size 0.4 mm
-S FT h.+B): 27 ft.
-- --- -- ----------------.............. ....... .....
Figure 4.13: Indialantic Project: erosion rate sensitivity, +2 to 5 ft/yr.
400000 O 300000 ) 200000
-- -- 7 -. ----
12 2 36 '18 60
MONTHS RFTER PROJECT
Drag scraper operations began in 1963 and continued through the late sixties with little beneficial effects. Borrow pits were located very near shore and the sand simply refilled within a few months (Walton, 1976). The first major nourishment efforts with offshore sediment were begun in 1973 and have continued with the most recent occurring in 1987. Table 4.2.4 summarizes the different projects by listing volume placements and length segments all within a large overall project area.
Sediment size from the offshore borrow area is small in comparison to the native sand with a median size of 0.12 0.15 mm. This sediment size is significantly smaller than all other project borrow materials analyzed in this study. Excess fine material is quickly worked off the shoreface and transported away from the fill area. A k factor of 1.5 was input along with an h. + B of 27 feet. CDN wave data interpolated between Vero Beach and West Palm Stations in conjunction with CERC data at West Palm result in an effective wave height of 2.26 ft and an average period of 7.5 seconds. Average background erosion from DNR over the project study limits show an average of 2.61 feet of erosion per year with significantly higher erosion in the northern segments of the Jupiter coastline. Quoted rates by Stauble and others range from 0 to 8 ft/yr. Input for the model will be the DNR reported rates varying with distance within the project construction limits (some 35,000 ft.), and associated with the model grid points.
The complexity of the resulting nourishment planforms, along with actual volumes placed in particular grid elements created a formidable system to analyze. With multiple nourishment events, of which Jupiter Island is a prime example, analytical methods are inadequate. Table 4.2.4 highlights the history of nourishment events which served as model input.
Figure 4.15 presents model results for the planform evolution. The value of using a higher quality borrow sediment can be illustrated in reference to Figure 4.15 and analysis of the 1981 monitoring profiles (Strock 1981). Twenty six profiles were reported with volume calculations. Profiles 2 through 24 showed a significant variance between the depth of the
Table 4.2: Jupiter Island Nourishment Summary. Signifies Placement in Multiple Segments
Multiple Reported Volume Per Year
1973 profiles and the monitoring depths of 1981. Nineteen of the profiles did not reach adequate closure for effective and meaningful comparison. The twenty three profiles taken in 1981 showed average shoaling of 1.7 ft over depths of 10 to 14 ft indicating large volumes of sediment had shifted offshore. This could conceivably result in an underestimate of 1,500,000 to 2,000,000 CY which lay offshore beyond the survey limits. Shown on Figure 4.15 is the associated range of possible error for the 1981 survey and its proximity to the model predicted curve.
Projects of long shoreline length show significant sensitivity to the erosion rate parameter input. Figure 4.16 plots model predictions based on uniform rates of 0, 5, and 8 feet of erosion per year in comparison to the DNR fluctuating rate (average of 2.6 ft/yr). The variation of predicted cubic yards remaining in the project limits at 15 years is 4.25 million CY. From the DNR data run to the maximum 8 ft/yr erosion the variance is approximately 2.75 million CY. This large effect on total erosion from the ambient rate is due to the 35,000 ft. project area. At 2.6 ft/yr, the erosion would equate to 91,000 CY/yr. An opportunity to study the effects of the background erosion rate as well as a chance to prove increased longevity from improved sediment characteristics is lost due to the lack of frequent and accurate profile surveys.
DATE QUANTITY (CY) LENGTH (ft.)
1973 2,519,362 17,821 1974 969,400 9,200 1977 267,000 2,588 213,066 3,600 1978 847,223 7,650 1983 594,000 5,850 406,000 3,150
1987 375,000 3,125
: o a o o. a
-O mQ '0 0
Figure 4.14: Jupiter Island: project locations for multiple nourishments. (Modified from Aubrey, 1988).
JUPITER ISLAND PROJECT: MODEL PREDICTION
Project completion date Initial design volume CY
1974 969,400 1977 480,066 1978 847,223 1983 1,000,000 1987 2,231,000
Effective wave height 2.28 ft. Sediment size 0.12 0.15 mm (h.+ B): 27 ft.
6000000 5500000 5000000 4500000
4000000 3500000 3000000
2500000 2000000 1500000 1000000
12 24 316
LI8 60 72 oil 9'6 io o 120 132 111 156 MONTHS AFTER INITIAL PLACEMENT
Figure 4.15: Jupiter Island: model volumes and indication of possible survey variance.
JUPITER ISLAND: EROSION RRTE SENSITIVITY
-2.6 Ft DNR
............. ... F tI
-- -5 Ft
...--... --8 Ft
I IroJe-t copletion diiit inMtiJTTIn volume (CY)
197 2,619,362 1974 969,400 1977 480,066 1978 847,223 1983 1,000,000 1987 2,231,000
Effective wave height 2.28 t.
Sediment glue : 0.12- 0.15 mm (h.+ B): 27ft.
Z (i5OIJ0l a:
2 '000000 -a:J
2000000I 500(00 10
1977 1978 1983 1987 48 A 84 9 108 120 132 1 4 1 a 0 MONTHS AFTER INITIAL PLACEMENT
Figure 4.16: Jupiter Island: erosion rate sensitivity, 0 to 8 ft/yr range.
4.2.5 Ft. Pierce
Fort Pierce is located on Florida's east coast in St. Lucie County, approximately 70 miles south of Cape Canaveral (figure 4.17). Ft. Pierce inlet was cut and jetties built in 1921 with jetty reconstruction in 1926-1927. Beach erosion just south of the inlet has plagued the region since the 1930's with an average recession of 4.3 ft/yr (U.S. Army Corps of Engineers, 1984). The nourishment area consisted of a segment of beach immediately south of Ft. Pierce Inlet's southern jetty. The project construction lasted one year from May 1970 through May 1971 and consisted of 718,000 cubic yards of sand placed on 1.2 miles of shoreline.
An interesting placement method was used here and consisted of an experimental underwater hydraulic dredge which had never been attempted previously. Problems forced the return to normal dredging practice after the first 60,000 cubic yards was pumped. A floating hydraulic pipeline dredge was used for the remainder of the sediment operating at a capacity of 20,000 CY per day. There is no data for the sediment characteristics of the borrow material.
Input for the data will simply be the assumption of .77 for the k value. Native material was in the range of .15 to .25 mm median grain diameter. The average wave height at this location is interpolated from the Vero Beach and West Palm Beach stations and resulted in an average wave height of 2.39 ft and an average period of 8.23 seconds. Background erosion rates were taken from St. Lucie County data compiled by DNR. The background rate in St. Lucie County is calculated for two time periods. The first is the 17 year period immediately following inlet construction, 1928-1945. The second period is a span of 22 years, 1945-1967. Figure 4.18 illustrates the erosion rate for the entire county shoreline. The location of the inlet at approximately monument R-34 is highlighted by the accretional area north, and the associated erosional area south of the inlet jetties. Recovery is evident farther downshore. The rates for the later time period are applied for model input to allow for the sharp change associated with the years closest to the inlet construction to equilibrate. The average erosion
rate over the beach nourishment area is 6.2 ft/yr (average of approximately 40,000 CY/yr of erosion). Of the 3.2 million CY dredged for inlet navigation improvements from 1930-1985, 2.7 million CY were dumped at sea, (Dean and O'Brien, 1987), and a mere 33,000 CY onto the beach south of the inlet.
Post construction profiling was carried out by the U.S. Army Corps of Engineers, (USAE), and was summarized in a four year study (USAE, 1975). Five post construction surveys covered this period and are compared to a pre-nourished profile. Methods employed preclude the use of the majority of the survey results. The pre and post construction surveys extended only 500 ft seaward of the baseline. Report analysis is suspect due to the fact that the later surveys extend to 1,000 and then to 1,500 ft off the baseline and direct comparisons are made to the 500 ft surveys. However, the volume change over a two year period, 1973-1975, is apparently of similar origin and extent. The volumes are calculated out to 10,000 ft south of the inlet or 3,500 ft past the nourishment limit.
Due to the absence of reliable data, a comparison is made between two boundary condition applications. The first application assumes normal wave approach and a grid network similar to case 2 (chapter 2). The St. Lucie erosion data discussed above provide the ambient erosion rate information. The second assumes QEP =0.0 and calculation is made of the associated shoreline angle downdrift of the jetty (case 3, chapter 2). A littoral transport rate of approximately 200,000 CY/yr acts at this area with a reported range of 140,000 CY/yr to 225,000 CY/yr (Dean and O'Brien, 1987). The resulting approach angle is 2.28 degrees. Figure 4.19 illustrates the predictions of the resulting volume. The plots include a placement of 33,000 CY dredged from the inlet and placed on south beach in 1974. Results of the different boundary condition applications shown in figure 4.19 reflect a large difference in performance prediction. The first method is suspect in boundary application, while method 2 agrees well with the analytical prediction until the introduction of the miscellaneous sediment at the three year time level.
The report of the two year period of erosion, (1973-1975), listed a loss of 222,000 CY out to 1,500 ft from the baseline while the model predicted 220,000 CY. The relative change is in good agreement. However, a figure of 27% is reported for the volume lost during the first 48 months, while model prediction results in 75%. Any conclusion is suspect due to the method of data collection utilized in the monitoring effort.
The Hillsboro Beach project is located in northern Broward County approximately 45 miles north of Miami. The project site lies within a 27,000 ft. stretch of coast bounded north by Boca Raton Inlet and south by Hillsboro Inlet (figure 4.20). This is a small volume and length project compared to the others included in this study. Approximately 385,400 cubic yards of sediment were dredged from offshore sites and pumped over a project length of approximately 5000 feet. Fill operations were initiated on August 16, 1972 and completed September 18, 1972.
Basically, the project area was divided into 12 cells alongshore, each 400 ft. wide along the project baseline. Strock reports (1973, 1974, 1975) present the data in volume per lineal foot per quadrant, based on establishment of 18 profile lines covering the project area and adjacent areas. Borrow sediment median grain diameter is .6 mm relating to a k of 0.55. Sediment studies performed in 1974, two years post nourishment revealed a beach composite sand size smaller than the borrow material, .4 mm. Two incidental sediment additions occurred after the initial nourishment in 1972. Hurricane Gilda (1973) introduced pproximately 4,000 CY of material stockpiled near the northern project limit. Also, 16,000 CY were placed in the northern project reaches at Port De Mar (Strock, 1975). Background erosion rates are determined by DNR survey reports over a 42 year period (1928-1970). For all Broward County Projects, wave height and period calculations were determined by interpolation between the West Palm and Miami CDN stations and CERC data from West Palm Beach. At Hillsboro, the average effective wave height is 1.87 ft and average period 6.3 seconds.
Figure 4.17: Ft. Pierce: project and inlet location.
- --- --1 Mile
ST.LUCIE BACKGROUND EHOSION RATES
SFt. Pierce Inlet
. \. / \
10 20 30 10 50
DNR MONUMENT NUMBER
Figure 4.18: Background erosion rates for St. Lucie County shorelines.
~ \ I\ j1
--. ..----70 80 9
FORT PIERCE VOLUME PR1 ICT I[NS
MODEL ME (HOD I ..................... MODEL MErHOD 2
.NnyL I I CAl
-1500011 -----------0 12
700000 650000 -600000 550000 500000450000 1100000350000 300000 2S000200000 150000 10000050000 0
2q 36 48 60 72 MNTIIS IFTER INIfA[. PI rFMENT
Figure 4.19: Ft. Pierce: model predicted volumetric changes.
N N -~
N N -
Project completion date: Initial design volume: Project length: Effective wave height : Sediment size mm (h. + B) 2t Qo Q2dr.
-Approach angle :
4/1971 718,000 CY 6,340 ft. 2.39 ft. ?? mm 27 ft. 10(),(00 CY
Table 4.3: Hillsboro Beach: Percent Remaining of Original Volume
MONTHS ANALYTICAL MODEL I SURVEY
4 84.1 80.1 78.9
12 71.5 72.7
15 67.4 66.2 74.9
24 55.7 59.5
28 50.5 59.4 69.1
36 40.1 53.1 48 26.9 45.8
The erosion rate decreases from a maximum of 2.23 ft/yr in the northern project area to 0.29 ft/yr in the southern reaches. This decrease could be connected with a long term buildup north of Hillsboro Inlet (12,000 ft south of the southern project limit). Boundary conditions of the infinite beach variety are assumed here with sufficient distances beyond the project in the model grid network. This assumption may be questionable due to the proximity of Hillsboro Inlet to the south. However, the project predictions only extend three years, not enough time for the spreading losses to reach 12,000 ft south.
Volume calculations from three monitoring efforts are reported and subsequently compared back to model predictions. The analytical and numerical results are seen in table 4.2.6 based on percent of original volume placed. The differences between the analytical and numerical results are due to the inability of the analytical method to incorporate the volume additions following the nourishment. This might not cause significant problems in a larger fill, yet these later additions amount to 6 percent of the original nourishment. Hillsboro is the second smallest volume project and significantly the shortest in length.
Figure 4.21 illustrates the two prediction attempts plotted along with the survey volumes. Model prediction at fifteen months showed a deviation of 32,000 CY and at 28 months of 35,000 CY in comparison with the survey results. Both of these are under 10% of the original nourishment volume. Sources of variance here include the unknown length of survey extent seaward. As illustrated in a previous section, this can introduce significant
Table 4.4: Hillsboro Beach: Cumulative Volumes of Erosion Components
NOURISHMENT BACKGROUND EROSION DIFFUSION EROSION
(months) (CY) (CY)
12 5,544 103,349 24 11,088 139,349 36 16,632 157,135 48 22,176 177,804
variances. Each of the projects studied in Broward county have been set up with a known h. value of 12 feet. Offshore of Broward County, the depth of closure is assumed to be 12 feet ((h. + B) =21)in contrast to 18 for a (h. + B) =27. This is due to a large expanse of rock further offshore from this depth and associated minimal sediment transport.
If the survey closure did not reach a minimum of 12 ft, volume misrepresentations can occur. This unknown must be evaluated properly in future projects to assure closure and more representative volume calculations. Table 4.2.6 separates the nourishment erosion into background rates and diffusion effects. The volumes listed are cumulative and can be evaluated similar to the Hollywood/Hallandale summary. The diffusion effects are seen to decelerate over time with 103,000 CY of sediment in the first 12 months reducing to 20,700 CY in the 12 month period from the third year to the fourth. The background effects remain constant at 5,500 CY of erosion per year within the project limits.
4.2.7 Pompano Beach
Pompano Beach is located in Broward County immediately south of Hillsboro Inlet approximately 30 miles north of Miami (see figure 4.22). Hillsboro Inlet, figure 4.23 is a natural tidal inlet improved with a northern jetty built on a natural reef in 1966. A gap in this section acts as a weir with a deposition basin immediately inside the jetty and an a dedicated dredge which pumps material from the basin to the south beach (Jones, 1977). It is estimated that prior to 1985 the bypassing rate averaged 60,000 to 80,000 CY and after improvements in the system increased the volume to approximately 100,000 CY per
Palm Beach County Broward County
North Construction Limit R07
Hillsboro Beach R12
South Construction Limit Pompano
North Pompano Beach
Pompano Beach SCale mile
Figure 4.20: Hillsboro Beach: nourishment project area
HILLSBOHO BEACH MODEL AND SURVEY 400000 OE
100000 Project completion date: 9/1972
Initial design volume 385,400 CY Project length : 5020 ft. Effective wave height : 1.87 ft. 50000 Sediment site 0.6 mm1 (A. + B) :21 ft.
0 r --- F
0 8 12 16 20 24 28 32 36 MONTHS AFTER INITIAL PLACEMENT
Figure 4.21: Hillsboro Beach: survey volumes and model prediction
year. Erosion problems were first addressed by an initial project undertaken from May to October 1970 covering three miles of coastline. The fill length of 16,800 ft was nourished with 1,033,000 CY of sediment from offshore borrow areas located in 30 to 65 ft depths and in three separate areas (Walton, 1977). The natural beach material was listed as a median grain size of .92 mm from surface sand samples due to a high shell content. Wave characteristics at this location interpolated from West Palm and Miami CDN data indicate an average yearly wave height of 1.8 feet with an average period of 5.8 seconds.
A major nourishment was again required due to severe erosional problems and was completed in August of 1983. The construction extended south from Hillsboro Inlet approximately 5.3 miles and included placement of 1,909,184 cubic yards of offshore sediment (CPE). The southern construction limit concluded immediately south of Lauderdale-by-theSea. The mean reported sediment size is 0.31 mm and a pre-construction beach median size of 0.59 mm. The resulting k factor is 1.2.
The major complication in predicting performance of the project and erosion/accretion of the area in general is due to the introduction of sand bypassing material at irregular times and often without record. DNR erosion data for this section of Broward County is unreliable due to the unaccountable volume introduction. Case 3 from chapter 2 is followed in model application. For comparison, the results presented include two approaches. Both include grid setup immediately south of a jetty, and an ambient transport rate, Q,, of 200,000 CY/yr. The first approach assumes a zero rate of sediment bypassing, QBP =0.0, and the introduction to the system of the bypassing quantity, 70,000 CY/yr before 1985 and 100,000 CY/yr after. This is treated by implementing a "nourishment" event each 6 months of half of the yearly bypass rate in the closest cell to the jetty. Q, QEP =200,000 CY/yr and the resulting wave approach angle is 5.25 degrees. Alternatively, the second approach assumes a bypassing rate of 100,000 CY/yr in the calculation of the defensive posture and results in (/6 ab) =1.21 degrees. Essentially the methodology is sound in both cases and the results are shown in figure 4.24. Model 12N stands for the "nourishment" approach,
the first discussed above. The timing of the increased bypassing amount occurs after 24 months of evolution resulting in the abrupt slope change. The second approach follows the more strict interpretation of the boundary condition. Analytical results are also plotted on figure 4.24 and are shown to mask the second approach.
Monitoring of the second project included 28 construction limit profile surveys as well as two control profiles south of the project, all utilizing DNR monuments. The 12 month survey, performed by CPE reflects an exact correlation with the second modeling case as well as the analytic prediction. A loss of 159,000 CY is predicted and realized. A five year survey performed by the County from the baseline to the 12 ft contour shows a cumulative erosion of only 82,700 CY. Method two predicts a loss of approximately 650,000 CY. The survey volume is suspect due to the time interval and minimal loss. Without other survey information, the success of the 12 month prediction can not be confirmed.
The Hollywood/Hallandale beach area is located in the southern reaches of Broward County approximately 15 miles north of Miami (figure 4.25). The effects of Port Everglades Inlet, 3 miles to the north provide the dominant causes of an erosion problem which has existed for many years. The inlet was opened in 1928 and training for navigation was completed in 1931. Background erosion rates drawn from DNR survey data over a 37 year period reflect variable erosional and accretional regions in part due to the movement and subsequent closure of Dania Inlet by 1944. The inlet migrated 5,000 ft from 1927 to 1936 and was located at that time 10,000 ft south of Port Everglades Inlet. The remaining data do not allow a calculation of change rate over a time period of anything greater than nine years. From a 1978 county report, the Army Corps of Engineers lists a rate of 2 ft/yr for an average yearly ambient shoreline recession. In a 1972 report on Port Everglades Harbor, a 5 feet per year erosion rate was listed for the shoreline south of the inlet which is again 3 miles to the north. The historical rate of 2 ft/yr from the county summary report is used for model input. The county report as well as the monitoring results were obtained from
Figure 4.22: Pompano Beach: project locations for 1970 and 1983 beach fills
Scale in Feet
\\ .North Jetty Constructed In 1930
Impoundment\\ Basin Weir Section
South Jetty Constructed In 1952 & Modified In 1964
North Jetty Final Construction 1965
Figure 4.23: Hillsboro Inlet: jetty construction history and impoundment basin location
PUMPrNU BEFICH MUDEL HNH SURVEY T
Project c0inpleLion date:- 8j8 Initial design volunme 1,909,184 CY Project ksngth : 28,00 ft. Elective wave height 1.83 ft. Sediment size 0.31 mn (h. + B) : 21 ft. Q.- Q01: 100,000 CY Approach angle 2.62 deg._ I
-f 8000000 Z I
600000 400000 200000 0
OP'= 100,000 CY
- --- -- -- 111f___ iIf
-MOIJEL I 20
.. I1IEI I N
11N1.II i F IE [ LACF1M[NI
Figure 4.24: Pompano Beach: model predictions and survey results
70,000 CY 12 24
Coastal Planning & Engineering. The average significant wave height acting in this area from interpolated CDN data is 1.62 ft with an average period of 5.56 seconds. Depth of closure is assumed equal to 21 ft for all projects in Broward County.
A high volume, yet short length, project was completed in September of 1971 with 360,308 CY of sand/shell placed over 4,000 ft with sediment in the range of .2 to .46 mm median diameter. No monitoring studies were performed on this original project resulting in a lack of understanding of project performance. By the late 1970's, the erosion problem was severe over a much greater length of coastline. To combat this problem, the two cities chose beach nourishment for a long stretch of beach on the southern portion of the county.
This nourishment project was begun July 31, 1979 over a project length of over five miles, 27,760 ft The southern construction limit lies on the county line between Dade and Broward Counties. The initial plan called for sediment to be trucked into the area; however, adequate supplies of suitable material were found offshore. Seven borrow areas eventually contributed to the project, all located between 5,000 and 10,000 ft directly offshore of the project limits (Suboceanic Consultants, 1980). The sediment was noted as "somewhat smaller" than the natural sediment. Adequate analysis of the native sand resulted in a median grain size of 0.35 mm with good sorting. Overfill ratios were determined for each of the borrow composites and ranged from 1.00 to 1.30 with an average of 1.09. This shows the slightly finer borrow material in comparison to the natural beach. A k factor minimum was thus set at 1.05 and after further investigation, a final factor of k = 1.15 was used as model input.
The original design volume for the fill included the following: design fill, loss of fines, and a five year advance nourishment. The resulting volume was 2,035,500 cubic yards of material. Dredging was completed on November 14, 1979 by the cumulative work of three dredges to avoid the associated delays due to rough weather of the coming winter. The pumped volume was approximately 97 percent of the design, a placement of 1,980,685 cubic yards. The distribution across the planform was far from uniform and is approximated in
Table 4.5: Hollywood/Hallandale: Profile Volume Changes With Time Between Survey Events
DATE MONTHS AFTER VOLUMETRIC CHANGE (CY) INITIAL PLACEMENT 11/79 0
the model application. This ranged from a 125 ft extension of dry beach to a mere 30 ft advance.
Monitoring was accomplished by profiling at 1,000 ft intervals after completion of construction from the upland limit of fill to the intersection of the offshore depth and the pre-construction survey profiles. Six months after completion, a volume of 1,804,806 cubic yards was calculated as remaining in the project area. This relates to a loss of 9 percent of the original volume. Further profiling was accomplished and is summarized in table 4.2.8. Volumetric change in CY is listed between the survey dates and is not cumulative. For example, 100,445 CY of material left the project area between 6 and 20 months following completion. The resulting volumes found provide the survey input for comparison to the model prediction.
In the numerical application, this area is treated as an open coast project with no direct influence from coastal features (see case 1 discussion earlier. A projection for the following 7 year period is included and shown in figure 4.26 along with the survey volumes. A good correlation results between the model prediction and survey volumes, particularly through
36 months. During this period the largest deviation was approximately 40,000 CY with very close agreement throughout. The average deviation through this period is 29,053 CY in contrast to the average deviation for the entire 6 years of 196,407 CY. The ability to predict within 3 years at an error of 29,000 CY is considered highly successful. Correlation decreases with time past the 3 year period and falls far off with the last survey at 74 months, showing a difference in volume of 424,000 CY. Also included in figure 4.26 are the analytical results through the same time period. Very close agreement between the model and analytic predictions is illustrated through the six year study. The deviation at 74 months is 45,384 CY.
Sensitivity to the background erosion rate is presented in figure 4.27. The 2 ft/yr rate was used in the results discussed above and is compared here to an increase to 4 ft/yr. A significant improvement resulted in prediction success. With the 4 ft/yr rate, the standard deviation decreases to 126,235 CY from 196,407 CY. Therefore, a simple adjustment of 2 ft in erosion along the project baseline improves the prediction "confidence" by 70,000 CY. This improvement is significant in areas where the erosion rate is undefined. An interesting approach in analysis is to differentiate between the erosion losses due to background erosion and those due to spreading out losses. Table 4.2.8 lists the percentages of each along with the total erosion loss in cubic yards. As expected it is shown that the diffusion loss percentage will decrease with time for the perturbation of the shoreline is spreading out and erosion is acting on a straighter and longer shoreline. The associated erosion percentage increase is misleading. This loss is constant on a yearly basis and is 65,333 CY at Hollywood/Hallandale yet becomes greater than 50% of the total erosion after only 4 years.
4.2.9 Treasure Island
Treasure Island is located in Pinellas County on a Gulf Coast west of Tampa Bay. The 3.5 mile long barrier island is typical of many west coast barrier island configurations and is bounded to the north by John's Pass and Blind Pass to the south (figure 4.28). The area has documented erosion from the mid 1800's to the present; however, the presence of the
Dania Cut Off CanalDaniaz
Nautical 1 1/2
Figure 4.25: Hollywood/Hallandale: project and borrow area locations.
HtOLLYWOOD/HALLANDAl E MODEL AND SURVEY
- 2I 1
12 2tJ 36
2000000 1 800000
4 00000 200000 .
-_ MODEL ...................... ANALYTIC L
--. SUV MT
Figure 4.26: Hollywood/Hallandale: survey, analytical, and model volume results
L Project Iompletion date: 11/1979
Initial design volume: 1,980,685 CY
Project length : -27,760 ft.
Effective wave height 1.62 ft.
Sediment size :.0.35 mm
._+_B) 2 ft._
48 60 72 aq
MONTHS AFTER PLACEMENT
HOLLYWOOD/HfILLFANDRLE MODEL HND SURVEY
Project completion date: 111979 Initial design volume 1,980,685 CY Project length : 27,760 ft. Effective wave height 1.62 ft. Sediment size :0.35 mmn
fhl_+ BJ_: __ 21ft .... ..
600000 1000000 Lii
S000 '400000 200(0000
-- -2 F I [00O .----........-------------------------------. 4 F Eno
1 ~ ---- 5 1VI T
Figure 4.27: Hollywood/Hallandale: sensitivity to background erosion rate, 2 ft/yr and 4 ft/yr.
12 214 36 18
MONTIIS AFTEB PLACEMENT
Table 4.6: Cumulative Component Percentage of Total Erosion ELAPSED TIME TOTAL EROSION PERCENTAGE
(months) (CY) background jj spreading
3 87,579 19% 81% 6 130,584 25 75 12 191,199 34 66 24 312,699 42 58 36 418,684 47 53 48 516,529 51 49 60 609,332 54 46
two passes causes significant problems in determining an erosion rate with any confidence. Wave data from the Clearwater location in the Gulf indicate an average wave height of 1.28 feet.
Initial restoration began with placement of emergency fill on the project site of 120,000 CY following Hurricane Gladys in October of 1968 is the first documented nourishment attempt. This was followed with 673,000 CY placed from April to July of 1969. This first project thus consisted of a total of 783,000 CY over a beach segment of 1.7 miles. The northern limit of this project was more than a mile south of John's Pass which at the time of improvement had a small southern jetty. The fill material came from offshore and from the shoals of Blind Pass, with the majority from the offshore site immediately gulfward of the project, 86%. A median grain size of 0.24 mm was determined from a weighted average of the three project reports and relates to a k factor of 1.3. The assumption on the west coast is a (h. + B) value of 15 ft.
A different approach at boundary conditions for this location is required due to the termination of the barrier island at two passes. The northern end of the island is treated as an application of a case 2 boundary condition with a zero transport northward of the island termination. This assumption is based on the beach protruding in the northern reaches of the island and enclosing O'Brien's Lagoon (Mehta et al., 1976). This effectively causes a no flow condition. The southern end of the project, northern side of Blind Pass, was armoured
with a very short jetty and was improved with a jetty extension in 1976. Effectively the boundary condition assumed here is a sediment sink into the pass. This is accomplished by forcing the first grid cell to zero for the duration of the prediction.
The University of Florida performed the monitoring profiles shown in figure 4.28 on similar lines of the 1969 pre- construction survey performed by the Corps. Results from 15 and 20 months following placement are recorded and are shown along with model predictions in figure 4.29. The two model predictions are based on different background erosion rates. Hobson (1981) referenced a background erosion in the nourishment area of 27,000 m3/yr, equating to 35,300 CY/yr. This is approximated as an average of 3.84 ft/year of erosion. To study the sensitivity to this approximation, the rate was doubled and the results plotted alongside the plot labelled "3.84 FT ERO" in figure 4.29.
A volume remaining at 15 months of 546,000 CY is reported comparing to a prediction of 606,407 CY (3.84 ft run) and 579,740 CY (7.68 ft run). At 20 months, the survey reported 546,500 CY remaining compared to 583,271 CY and 547,715 CY for the two prediction results. The results are encouraging and could possibly indicate a higher background erosion rate than that previously reported. Uncertainties are introduced however due to the boundary conditions at the site. Six months after the 20 month survey, a renourishment of 76,000 CY placed sediment in an are just north of the original northern construction limits. A second renourishment was completed in 1972 placing 150,000 CY on a 1,400 ft stretch in the lower limits of the original nourishment area. The multiple nourishments are not included in the analysis due to the absence of any monitoring data after the 20 month report of 1971. The extension of the jetty in 1976 creates a boundary condition similar to either case 2 or case 3. Improved representation of boundary condition applications may enable an improved performance prediction.
4.2.10 Captiva Island
Captiva Island is a barrier island located off the west coast of Florida approximately 100 miles southeast of Tampa Bay (Figure 4.30). The Lee County island is bounded to
1971 Fill Area 1969 Fill Area 1972 Fill Area 0
Figure 4.28: Treasure Island: nourishment project areas and proximity to Blind Pass and John's Pass.
IFF1?[JHF- [ (d HNIH M1][I )NH] SlJHV[ Y
400000 LLI 350000
300000 i) o' o () CD
Project completion date: 7/1969 Initial design volume : 783,000 CY Project length : 9,000 ft. Effective wave height : 1.28 ft. Sediment size 0.24 mmin (h. 11i) 15 ft.
AFTEH PLACEMENT -F IEi
Figure 4.29: Treasure Island: survey volumes and model prediction through a two year period.
the north by Redfish Pass and to the south by Blind Pass. Captiva Island has experienced high rates of erosion throughout the years with documented erosion rates from as early as 1876. Redfish Pass was formed by a hurricane in 1926 and has remained relatively stable since that time unlike Blind Pass which has a history of closure and southern migration. The original nourishment project was completed in October, 1981 with a design volume of 765.00 CY. Actual placement volume was reportedly greater over the northern 10,000 ft. of coastline. A short terminal structure was built at the time of the 1981 nourishment. The original design for the project was based on an average annual sediment erosion rate of 5 CY/ft/year (Barnett 1988).
The inlet ebb tidal shoal of Redfish Pass provided high quality material and construction began in July, 1981. A reported median sediment size of Dso = .44 mm results in a k value of 0.82. Monitoring information was adequate; however, some problems in the use of these data exist due to nonstandard baseline survey locations. Wave details include an effective wave height of 1.44 ft. and an average period of 4.7 seconds. Background erosion rates from DNR range from 1 to 30 feet/year. A long term uniform rate of 3 ft/yr is used in the modeL
Near inlets such as Redfish Pass which do not have ideal structures, the boundary conditions are not clearly defined. Here, the large ebb tidal shoal provides a substantial sheltering effect for particular wave directions. In contrast, Redfish Pass can act as a total sediment sink. Thus, the appropriate boundary conditions to be used at this site are not evident. Basically, the boundary condition applied here is one of forcing a zero transport northward of the project limit. As noted previously, this is equivalent to an effective project length of twice the actual length, case 2 in figure 2.2. Table 4.2.10 presents the survey dates and cumulative volumetric changes (erosion (-)).
Figure 4.31 presents a plot of projected model cumulative volume losses against the ac=-,-i survey losses in the project area. The correlation through the first 12 months is outstanding with a deviation of less than 5,000 CY, or less than half of one percent of the
Table 4.7: Captiva Cumulative Volume Change Months after project Cumulative
completion Volumetric change (CY)
12 -68872 18 -120.560 41 -164,560 46 -109,260 52 -169,760 58 -196,760 65 -202,075
original volume. An over-prediction of loss by approximately 20,000 CY at 24 months is followed by a converging loss prediction. At 41 months. a survey calculated loss of 164.500 CY is comparable to a predicted loss of 164,000 CY. Following this high correlation, the survey of 46 months reflects a significant accretion in the profiles with a gain of over 50,000 CY. This could be explained by profile error due to a varying baseline or by a reflectim of storm recovery due to a short term event. However, the validity of the model is verified from this time forward ending in a difference of less than 25,000 CY at 65 months, less than 3% of the original placement volume. These results are certainly encouraging particularly for the application of 'case 2" project boundary conditions. The predicted volume loss remain a good approrimation throughout the life of the project. An overall standard deviation of 31,693 CY results from placement through 65 months. This reflects a prediction ability capable of error of less than 4% of the original volume.
The lack of accuracy in boundary condition application with the proximity of the notxhern project boundary to the inlet, introduces uncertainties. Also, a high reported shell content in the pumped sediment could significantly reduce the sediment parameter, k. A measure of wave height sensitivity, by increasing and decreasing the effective wave height by 20% resulted in a uniform 3% deviation in prediction results.
A major renourishment is currently underway at Captiva, extending further south than the 1981 southern construction limit.
i2 North Captiva Island 1500' 6
2000 R !A~~B8
Oagtlva---letana --Gulf Af Mexico
Scal In Mles
:i Limits of Beach Fili
'///// Limits of Borrow Site
1.000 0 1,000
Scale In Yards
Figure 4.30: Captiva Island beach nourishment ;-.jec: azd borrow area locations.
CFlPT IVF- P H ECT: M0-IEL.. &H iHHVL 3S
-. ..........- SURVE T
PIroject coii pleiion date: I/ 1Af Initial design volume 765,000 CY Effective wave height : 1.44 ft. Sediment size 0.44 mii (h. + B): 15 ft.
211 3b '10 MONTHS F0IlUN NING PR[1LC I
Figure 4.31: Captiva Island Model and Survey Losses
2()00000 150000111000 50000 -
5.1 Summary of Investigation
It is commonly accepted that beaches widened by nourishment projects can be effective in providing the desired defense to shoreline erosion and reduced jeopardy due to storms. One measure of project performance is the time history of volumetric retention. An effort to predict the resulting planform and associated volume changes has been accomplished here with varying results. To attempt to present an assessment of the success by a percentage would over simplify the problem; however, an encouraging degree of success was evident in most of the cases studied. The present level of prediction methodology is not without certain limitations, yet provides a very promising and needed approximation to shoreline evolution. Prediction success, sensitivity summary, limitations of approach, and recommendations for further work are briefly addressed below.
5.1.1 Prediction Analysis
Present numerical techniques can provide reasonable estimates of project evolution with proper care in wave height, sediment transport characterization, closure depth approximation, and background erosion introduction and evaluation. In a general evaluation, the results at seven locations reflect success in prediction. The results at Delray Beach, Hollywood/Hallandale, and Captiva are more striking due to the amount of data available for better comparison. At Delray, the prediction over 14 years and through three nourishments, resulted in a standard deviation of 242,000 CY or 6.6% of the cumulative nourishment volumes. This small percentage is very encouraging in a prediction sense particularly with such a long time period. The Captiva application with different boundary conditions re-
Table 5.1: Sensitivity of k Parameter Compared to Dean Relationship Value, k, PROJECT C nI [a L k =0.77
Delray Beach k, =1.3 .12 .21 14 Cape Canaveral ka =1.1 .18 .04 2 Hillsboro ka =0.55 .11 .16 3 Pompano Beach ka =1.2 .01 .01 1 Hollywood/Hallandale ka =1.15 .21 .23 6 Treasure Island ka =1.3 .05 .13 2 Captiva Island k. =0.82 .04 .04 7
suited in a standard deviation of 31, 700 CY or less than 4% of the original nourishment. Hollywood/Hallandale results showed a standard deviation over seven years of less than 200,000 CY or 9% of placed volume. Over the first three years of the project, the average deviation was only 29,053 CY, less than 2% of the original nourishment volume. This degree of prediction capability certainly reflect success in the model applications.
5.1.2 Parameter Sensitivity
The relatively promising results and insights into parameter influence seen in the various project applications establish the utility of the prediction schemes utilized here. The results could thus be optimized by introduction of certain parameters. In the majority of cases, the model predictions have been on the conservative side throughout the time history. This could conceivably confirm that linearization of the transport equation over-estimates shoreline evolution. Sensitivity of the processes to several parameters was analyzed and found significant in regard to background erosion, effective wave height, refraction application, and sediment sizes. Improvement in compilation of these data and increased accuracy in profiling is ultimately important in future prediction success.
The background erosion effect is prominent on long projects due to the resulting volumes and is shown at Jupiter Island to amount to 4.5 million CY difference in interpretation. Sensitivity of the sediment factor, k, is illustrated in table 5.1.2. The nature of the sediment parameter and the normalized standard deviation with the constant k = 0.77 is illustrated
in the table. A value of c = 0.0 reflects the best possible agreement, while a value of 1.0 the worst. These values are obtained by calculations of model output with available survey information in a normalized standard deviation format. Improvement in prediction by using a k factor of 0.77 resulted in only one case. This would suggest that the relationship utilized between sediment size and the k factor is a better measure of sediment characteristics than the constant 0.77. With improved confidence in each of the terms listed above, a wave height change is highly significant. This term in the diffusion coefficient is to the 5/2 power. Thus a change from a 2 ft Hb to a 2.5 ft wave climate results in a diffusivity term increase by a factor of 1.75. This effect suggests the need for a detailed analysis and introduction of a well established wave height properly converted to a breaking wave height. Changes introduced by refraction, diffraction, damping, and other factors cause wave climate changes that should be addressed by further research interests. A further analysis of the relative importance of the assumed boundary conditions is also needed. This is particularly true at inlets as seen in the comparison of methods at Pompano Beach and Fort Pierce. The adaptation of local conditions such as the sheltering provided by Cape Canaveral shoals could then be included in the analysis.
5.1.3 Application Limitations
Limitations in application of the methodology outlined here result from the basic assumptions as well as in the choice of analysis method. Realistic wave/bottom interaction through a more thorough refraction application would allow a more appropriate wave climate at individualized shorelines. This effect is more important in the vicinity to coastal structures and areas with significant contour variation. Other areas introducing limitations are the neglect of diffraction around structures and the effects of wave damping upon wave conditions near the shoreline. Altered responses would result from changes in the wave approach angle along the wave crest. For this study, the wave approach angle has already been assumed small for linearization of the differentiated sediment transport term. Wave approach and shoreline angles can be handled by nonlinear approaches solved both
implicitly and explicitly.
Both the analytic and linear solutions assume parallel and straight bottom contours as well as nearly normal wave attack to linearize the combined equation. With approach angles less than 10 degrees, model predictions from the nonlinear explicit solution reflect a variance of less than 3 percent with volume remaining calculations.
More severe limitations in prediction and evaluation of nourishment events are introduced by the lack of available and adequate data with which to calibrate and evaluate the methodology.
5.1.4 Nourishment Monitoring Importance
The initial phase of this study comprised the search and compilation of adequate data on beach nourishment projects. The examination of the various nourishment and renourishment projects documented the lack of uniform monitoring and reporting of relevant data. The task of analyzing, archiving, and the frequency of both in future projects should be more thorough and consistent. Prediction capability can only be evaluated through accurate monitoring surveys. This was illustrated in several of the projects included in this study. In particular, the questionable survey data at Fort Pierce and Pompano Beach, and the lack of profile closure and the limited number of surveys at Jupiter Island and others severely limit the performance evaluation of significant nourishment events.
Improvement in monitoring schemes and compilation of relevant data are essential to shoreline evolution prediction as well as individual project evaluation. Monitoring guidelines need to be developed and followed to provide a common data base and therefore more accurate means of prediction and comparison. The need is strongly indicated for improvement in the following areas:
pre and post project surveys
frequency of comparative profile surveys
* reference of profiles to a common baseline
extent beyond profile closure
comprehensive sediment analysis (native and borrow)
Stauble (1983) outlines a detailed plan for the above items. Systematic and comprehensive monitoring plans must also be adequately collect, compiled, and archived. These major limitations are outside the scope of this or any research and require changes in the coastal community as a whole.
5.2 Recommendations for Future Work
Further studies in this area of coastal engineering should be directed specifically toward several areas including; increased parameter sensitivity analysis, boundary condition application, increased wave parameter analysis, and approaches beyond the scope of a one dimensional model. Variations in wave characteristics, due to seasonal changes and storm activity, cause an immediate on- offshore movement of sand. Unless the event is unusually severe, the profile will resume its original shape through cross shore sediment movement. However, imbalances in the longshore transport rate create erosional activity on a more gradual, and permanent, scale. In shoreline climates where cross shore effects are minimal compared to longshore processes, the one dimensional model is a good approximation and results in reasonable evolution prediction.
The addition of diffraction effects and the variability of wave height and angle on the shoreline would require a nonlinear model analysis. One step further, application of an "n" line model would direct specific attention toward important parameter sensitivity and could conceivably more truly represent the planform and profile evolution. A two-line theory can be applied after further research in the distribution of longshore transport across the surf zone (longshore and on-offshore). CDN gages capable of direction, p-u-v gages, would allow further investigation into the nonlinear model.
Detailed analysis of the effects of sediment size toward predicted evolution is another important area which can affect project performance from the outset. The use of wave data
collected over the time frame of the prediction would result in "real time" application and effects of storm events could be analyzed.
From the outset, the desired focus was identified as the analysis of current prediction capabilities. Considering the variability of the project applications with associated boundary condition uncertainty and questionable monitoring data, predictive capability is generally good. It must be emphasized that only through increased attention to monitoring efforts will improvements in evaluation as well as prediction be made. Adequate compilation of project reports from standard monitoring methods is of utmost importance in many areas of coastal engineering. Enhanced performance prediction would surely benefit the design and construction aspects of nourishment as well as ultimately reduce the construction costs and permitting delays and possibly avoid or minimize adverse environmental effects.
Arthur V. Strock & Associates, Inc., "Town of Hillsboro Beach, Beach Restoration
Project, Follow-up Report No. 1," Deerfield Beach, Florida, 1973.
Arthur V. Strock & Associates, Inc., "Town of Hillsboro Beach, Beach Restoration
Project, Follow-up Report No. 2," Deerfield Beach, Florida, 1974.
Arthur V. Strock & Associates, Inc., "Town of Hillsboro Beach, Beach Restoration
Project, Follow-up Report No. 3," Deerfield Beach, Florida, 1975a.
Arthur V. Strock & Associates, Inc., "City of Delray Beach, Beach Restoration Project
Follow-up Report No.2, 16 Month Study," Deerfield Beach, Florida, May 1975b.
Arthur V. Strock & Associates, Inc., "Town of Jupiter Island Follow-up Study," Deerfield
Beach, Florida, 1981.
Aubrey, D.G., and N.M. Dekimpe, "Performance of Beach Nourishment at Jupiter Island,
Florida," Beach Preservation Technology First Annual National Conference Proceedings, Gainesville, Florida, March 1988.
Bagnold, R. A., "Beach and Nearshore Processes, Part I, Mechanics of Marine Sedimentation," The Sea, Vol. 3, Hill, M.N., Editor, Interscience, 1963.
Barnett, Michael R., and Richard W. Stevens, "Performance of Beach Restoration at
South Seas Plantation, Florida," Beach Preservation Technology First Annual National Conference Proceedings, Gainesville, Florida, March 1988.
Coastal Planning & Engineering, Inc., "City of Delray Beach Second Periodic Beach
Nourishment Project 27 Month Follow-up Study", Boca Raton, Florida, 1987.
Coastal Planning & Engineering, Inc., "Pompano Beach/Lauderdale-by-the-Sea Beach
Restoration Project, 12 Month Monitoring Report, Boca Raton, Florida, 1985.
CRC Standard Mathematical Tables, 27th Edition, Editor; Belger, William H., CRC
Press, Inc., Boca Raton, Florida, 1984.
Dean, R.G., "Engineering Design Principles,"
Short Course on Principles and Applications of Beach Nourishment, Beach Preservation Technology First Annual National Conference, Gainesville, Florida, March 1988.
Dean, R.G., and R.A. Dalrymple,Water Wave Mechanics for Engineers and Scientists,
Prentice-Hall, Inc., Englewood Cliffs, New Jersey, 1984.