EDITORIAL A Time to Look Back at Beach Replenishment THE EAST COAST BEACH REPLENISHMENT EXPERIENCE onstrates the danger of importing generalizations from one shoreline system to another. In addition, U.S. East Coast beach design usually involves the assumption that pre-project erosion rates (after some initial profile adjustment) will be a good measure of post-project erosion rates. With the single exception of Miami Beach, replenished beaches have eroded much faster (lv2 times to 12 times) than their natural counterparts. Not surprisingly, storms seem to be the major factor determining beach longevity on the East Coast. At the time ofthis writing several new or proposed beach replenishment projects on the U.S. East Coast are being designed (at significant cost to communities or the federal taxpayer) with the same assumptions and the same equations that have proved to be inaccurate in the past. For example, in Vero Beach, Florida, a joint Corps of Engineers/engineering consulting firm design assumes that, because the fill grain size is coarser than the native grain size, the rate of erosion of the replenished beach will be substantially less than the natural rate. Maybe so, but this has never happened on previously replenished East Coast beaches. In February of 1989, riding on wings of optimism about his community's new beach, the Mayor of Ocean City, Maryland, came down to Vero Beach, Florida, to extol the virtues of beach replenishment. However, shortly thereafter, within 6 months of emplacement, in the first winter storms of 1989 (March), much of his $12 million beach disappeared. In places, the Ocean City shoreline had retreated to its pre-project location. Figure 1 shows two cross sections of a typical beach replenishment project. The lower one, shown in true scale, gives some idea of the insignificance of the size of the sand body pumped up on the beach, relative to the shore- Along the East Coast barrier island shoreline of the U.S., more than 90 beaches have been replenished in the last three decades in 270 separate pumpings. We find, in a study published in this issue of the Journal ofcoastal Research, that in spite of this vast amount of experience, coastal engineers and geologists have a poor record predicting the durability of artificial beaches and that they are basing their calculations on beach design parameters which don't work. Furthermore, the principles of beach design remain untested because nobody has looked back to evaluate the success or failure of past projects. North of Florida, we find that, essentially, no replenished beach has lasted more than five years (without major nourishment), and yet several recently published predictions of expected beach durability range from 10 years to 40 years. We also find that the widely-held assumption of a direct relationship between grain size and beach durability may not be entirely valid. In our study, we cannot demonstrate that differences in "imported" sand grain size relative to "native" grain size (within the narrow range of sand sizes used on the East Coast) make any difference in beach durability when viewed on a regional scale. Ironically, the entrenched belief in grain size effect on replenished beach longevity may have resulted in unnecessary expenses for some communities in obtaining the "correct" sand. There also seems to be no good relationship between beach length and beach durability of East Coast beaches. However, in a parallel study, we have found that longer beaches have greater durability on the Gulf Coast. That the relationship holds on the Gulf Coast and not on the East Coast demiii
WRIGHTSVILLE BEACH DESIGN BEACH -L- -L ~11.1-~~~~._-~--'O=====:~=2=7=o:o:,===========I="mmGG""'O' TRUE SCALE Figure 1. Cross sections of the Wrightsville Beach, North Carolina replenishment project; upper diagram has 5x vertical exaggeration and the lower diagram is at true scale. The relative size of the replenishment projected is indicated by the slight protuberance above the junction of the sea level and sea floor lines. face. It is well to keep in mind the scale ofbeach replenishment projects, a point emphasized later in this editorial. Our study is not the final word. Future studies with more data based on better monitoring may significantly modify our conclusions. But our results and conclusions deserve careful consideration by those who design beaches or predict sand behavior. WHERE WE'VE GONE WRONG I believe that our failure to properly design beaches and predict beach life is due to several factors. These include (1) a strong belief by the engineering community in the theoretical approach to sand transport, (2) the failure to look back and test beach design principles in the real world, (3) the lack of real communication between design engineers on the one hand and coastal dynamicists and oceanographers on the other and (4) having storms as scapegoats for failed predictions. In my opinion, the biggest problem is that we have never looked back. The point is strikingly demonstrated at hurricane conferences and other professional meetings where the all-out effort of meteorologists to learn from past predictions stands in stark contrast to the lack of systematic hindsighting of beach nourishment projects by the coastal engineering community. Engineers and geologists have peeked back occasionally, but the general lack ofbeach monitoring has forced us to depend on studies ofsingle beaches to evaluate the success of our beach design parameters. The single-beach approach tends only to reinforce the accepted parameters (e.g. coarse-grained beaches disappear because of storms and fine-grained beaches disappear because of too fine grain size etc.). Only the iv
analysis of a large number of uniformly monitored beaches will allow us to improve beach design and to better serve the public. In reviewing the literature on the "engineering of sand" we have observed that there is virtually no communication between the scientists who study shoreface sediment transport processes (e.g. Don Swift) and coastal engineering practitioners. We also note that design engineers, who do the actual estimates of beach durability have, at best, limited communication with engineering scientists who do basic "surf zone science." For example, the studies at the Corps of Engineer's, Duck, North Carolina wave pier seem not to have permeated, even slightly, the practical world of replenished beach design. The lack of communication between engineering scientists and those who actually design projects is further illustrated in the latest edition (1984) of the U.S. Army Corps of Engineers' Shore Protection Manual (often considered the "bible" of coastal engineering). There are repeated cautionary statements in the manual; for example, "both models are simplistic descriptions of complex beach relationships" (p. 5-13). Something gets lost between the bible and the real world, and the models and concepts are applied with no recognition or mention to the public of their major weaknesses. Finally, the reason that the public accepts unexpectedly rapid beach loss and one reason we haven't looked back is because of the occurrence of storms. Our review of documents intended for public consumption shows a tendency to view premature beach loss in a storm as bad luck or an unavoidable accident rather than the predictive design failure which it actually is. WHY APPLICATION OF SAND TRANSPORT THEORY DOESN'T WORK Predicting the behavior of sand (the engineering of sand) in association with either hard structures or replenished beaches requires three tiers of assumptions. These are: (1) assumption of a wave and storm climate. (2) assumption of what the climate will do to sand movement and (3) projecting the first two assumptions, 10 to 50 years or more into the future. Assumption of a storm climate (#1) involves coming up with numbers that represent a large variety of conditions including storm waves of many fetches, durations and intensities. Waves from several storms may simultaneously interact and wind set up and set down will produce a variety of currents interacting with waves and other types of coastal and surf zone currents, including ripcurrents. In other words, we need an accurate holistic view of storm-driven coastal hydrodynamics. Assuming what the storm climate will do to the sediment (#2) is also complex. Sediment response will vary with grain size and sorting, presence or absence of shell lags, bedforms, degree of compaction, presence of organic films and mucus, water temperature, effect of burrowing and trailing organisms and many other small scale factors. Larger scale factors include shoreface, continental shelf, barrier island or coastal geomorphology. The time gap (repair interval) since the last storm is also important. These among other factors make it impossible to bridge the gap between the small-scale physics of water-sediment interaction that is somewhat understood, and the kilometer and the years scale behavior of replenished beaches that is the desired end point. Projection ofsand behavior and storm climate assumptions into the future (#3) is made particularly difficult by the occurrence of unpredictable bathymetric changes which change wave refraction patterns and impact on both storm climate and sediment transport. In summary, you cannot mathematically characterize a system that is not understood. THE DUTCH APPROACH TO BEACH REPLENISHMENT The Dutch have approached coastal management in general and beach replenishment in particular, in a most organized and systematic fashion. According to papers by and personal communications with Hendrik J. Verhagen, the Dutch generally do not use mathematical techniques for basic design of replenished beaches. Required sand volumes are determined by calculating the rate of sand volume loss from the coastal reach in question and assuming that post and preproject erosion rates will be essentially the same over the time span of a desired beach life. A 20% overfill factor is often applied to account for events such as profile adjustment and lateral losses. Other parameters such as beach length, grain-size, density and method of placing sand are considered to be subordinate to the pre-pro- v
ject erosions rate in predicting life span of artificial beaches. Such an approach has the distinct advantage of not depending upon any detailed knowledge of shoreface processes, local wave climate, fill factors, renourishment factors, closure depths, etc. It leaves for future generations the task of detailed study of shoreface processes and the processes which affect loss of replenished beaches. The assumption that pre and post-project erosion rates are nearly the same may work for the Dutch Coast but, as previously mentioned, the two are not the same on the V.S. East Coast. WHAT THE U.S. PUBLIC HEARS After a beach disappears, politicians, consultants and Corps officials try to put the loss in the best light possible. But in so doing a number of insupportable statements are typically made. These are paraphrased here along with my opinion of their validity. It might be noted that each of these statements points to a direction of much needed research. (1) The replenished beach will recover during fair weather. The actual statement published by Ocean City, Maryland officials was: "During the winter beginning in 1988, some ofthe new sand will move offshore and the Atlantic will appear to be reclaiming the new beach. Not so! The sand will settle in a nearby offshore area to stabilize the beach at a gradual slope and sand will be back in the spring." FALSE. Replenished beaches do not recover from storms like natural beaches. Storm recovery, if any, of artificial beaches involves only a small percentage of the lost volume. (2) While the sand lasted, the replenished beach served a protective function for shorefront development during the storm. TRUE. ceeding storm; the bigger the storm, the bigger the step. (4) The lost sand has simply moved offshore where it continues to serve a protective function for the community. PROBABLY FALSE. There are few data indicating where the sand goes and no data indicating whether the sand continues to affect wave climate at the shoreline. Vibracoring off Wrightsville Beach, North Carolina, by Stan Riggs indicated that the sand lost from this oft-replenished beach was widely dispersed in a thin layer over the inner shelf. Not surprisingly, we find that the public is not impressed with the contention that the sand is still "functioning" underwater. They can't see it and they want a recreational beach above water. As an example, at the recent Coastal Zone '89 meeting, the Myrtle Beach, South Carolina, replenishment project was characterized as being a great success. Although there has been significant shoreline retreat, 100% of the replenished sand can be accounted for, either on the subaerial beach or in the active offshore profile zone. I question this judgement of success. The communities main objective was to improve their recreational beaches and sand residing offshore doesn't help in that regard. As to whether or not the offshore sand is playing some other useful role or whether it will stay there much longer to play that role has not been determined. (5) The next beach nourishment should last longer because the sand lost from the previous beach reduces the need for profile adjustment of the new beach. FALSE. There are no indications of slowing down of beach loss with multiple replenishments over the last 25 years on heavily-replenished beaches in North Carolina, New Jersey and Florida. (6) The beach was lost due to unusual and/or unexpected storm activity. MISLEADING. Whether a storm is unusual is a matter of definition. All beaches have storms and their probability of occurrence should be part of beach design as well as part of the information given to the taxpaying public. No storms are "unexpected." CONCLUSIONS I believe that predicting the behavior of nearshore sand on the V.S. East Coast in the time (3) This loss of sand represents an initial high rate ofsand loss as the profile of the replenished beach is adjusted. Once the profile is adjusted, the erosion rate will slow down. SOMETIMES TRUE. This erosion rate plateau (whose nature and existence needs documentation) is probably not an important factor in beach durability. A more general picture of replenished beach erosion is a series of step-like losses with each sucvi
frame and with the accuracy claimed in many government and consulting documents we have read, is impossible. More important, attempting to predict sand behavior in the the same fashion that we predict the durability of stone and steel structures is a fundamental misreading of the sporadic, irregular and even chaotic nature of the "events" which control sand behavior. In other words, because storms are a major control of sand behavior and because we can't predict when the next "big one" will hit, it makes no sense at all to predict beach life span and express it as a single number. With regard to beach replenishment, I conclude that we should: (1) Make durability predictions in terms of probability ranges of beach lives, e.g. this initial replenishment will last five years plus or minus four years and the cost will range between 10 & 50 million dollars to keep a beach in place for 10 years, etc. (2) Until we achieve a better understanding of the engineering of sand, base such predictions on the durability experience on nearby beaches or on the behavior of previous beach replenishment projects on the beach in question. (3) Monitor all beaches planned for replenishment in the future to provide a basis for understanding the principles of replenished beach design. Do not allow replenishment unless at least a rudimentary monitoring program is already in place. Simply periodically measuring subaerial beach width would provide a very important data base for future design of replenished beaches. (4) Make the public aware that beach replenishment is costly and temporary, that it requires a long term commitment to work and that the first beach replenishment project is an experiment. Virginia Beach, Virginia, is an example of a community that has successfully conducted a long and continuous program of beach replenishment. (5) Begin to conduct research on the sand transport systems on the time and space scales of beach nourishment, i.e. years and kilometers. My thanks to Don Swift, Kathie Dixon and Tonya Clayton for reading and criticizing earlier versions of the manuscript. Orrin H. Pilkey Program for the Study of Developed Shorelines Department of Geology Duke University Durham, North Carolina vii
A Word Abollt Melnbership The Coastal Education & Research Foundation [CERFJ is a nonprofit corporation dedicated to the advancement of the coastal sciences. The Foundation is dedicated to the multi-disciplinary study of the complex problems of coastal environments. The purpose of CERF is to help translate and interpret coastal issues for the public and to assist professional research and public information programs. The Foundation specifically supports and encourages field and laboratory studies on a local, national, and international level. Through the media ofscientific publication, television, and radio, CERFbrings accurate information to the public and coastal specialists on all aspects of coastal issues in an effort to maintain or improve the quality of shoreline resources. CERF concentrates its efforts on significant problems such as maintenance of quality (potable) water with an adequate supply, and hazards associated with potential beach erosion, flooding, and susceptibility of developed shorelines to storm surge and wave attack. CERF membership is available to individuals, institutions, and corporations, that support the aims of the foundation through personal and group efforts or by donations. For membership information contact: Dr. Charles W Finkl, Jnr. Editor-in-Chief 4310 NE 25th Avenue Fort Lauderdale, FL 33308 USA (305) 565-1051