PART 3 CONTRIBUTED PAPERS (Contd.)

SCS/82/CFE/CP-2

OBSERVATIONS ON ENGINEERING DESIGN, LAYOUT AND CONSTRUCTION OF COASTAL FISHPONDS IN THE SOUTHEAST ASIAN REGION ¹

by

R.G. Hechanova²

1. INTRODUCTION

Observations of the writer on the state of engineering design, layout and construction of coastal fishponds in the Southeast Asian region and the evaluation and suggested solutions to the engineering problems encountered in some of the farms in the region are presented.

A case study for the renovation and improvement of a special project in North Sumatra reflects the need for careful planning, extensive data collection and repeated surveys to determine the physical and chemical characteristics of the mangrove swamp site prior to construction. The consequential failures, one of which is that of piping, bear out the need for such a careful survey.

1.1 Time and place of observations

Under the writer's terms of reference during consultancy assignments with FAO/UNDP projects, visits were made to various brackishwater (as well as freshwater) fishery stations and fishpond centres in the following places:

Indonesia

Jepara, Central Java
Semarang, Central Java
Prigi, East Java
Kalianyar, Bangil, East Java
Probolinggo, East Java
Bedagai Demonstration Ponds, North Sumatra
Babaian Demonstration Ponds, North Sumatra
Air Joman, Asahan District, North Sumatra
Perupuk, Asahan District, North Sumatra
Sialang Buah, North Sumatra

Malaysia

Ban Merbok proposed sites, Kedah
MAJUIKAN Siakap Farm, Kuala Sala
Proposed project site for MAJUIKAN in Sungai Tumboh
Gelang Patah Brackishwater Research Station. Gelang Patah, Johore

Philippines

Various sites and projects during the writer's consultancy period from 1977 to February 1982.

Thailand

Brackishwater and freshwater fisheries stations in the following locations were visited.

Chachoengsao
Samut Sakhorn
Phuket
Songkhla
Satul

Vietnam

Halong Bay Brackishwater Fishfarm
Haiphong brackishwater fishfarm of the Shrimp Culture Research Institute, Haiphong
Proposed brackishwater sites for Brackishwater Shrimp Hatchery and culture ponds in Qui Nhon province

2. OBSERVATIONS AND DEFICIENCIES NOTED

2.1 Design and general layout used in farms

The use of land for aquaculture sometimes results in some negative effects on wildlife and adverse environmental influences on the life of the people in the area near the farms. For example, it has been felt that the benefits of the aquaculture industry have been partly offset by flooding as a result of changes made on the regimen of the river.

In a survey of the fishponds in the flood plains in Central Luzon, Philippines, fishpond layouts present a conflict with other activities in better land use. Pond developers and owners saw little of the disastrous flood effects of the constriction of the stream flow, and constructed their dikes close to the river banks.

Illegal construction of fishpond dikes along the different tributaries of Pampanga River constricted the main water-ways. Expansion of areas for pond operation from both sides of the river banks made the waterways narrower.

Generally, in most fishfarms fed with riverine waters in the region, gates are often built close to the river intake, without providing a canal stretch which could be utilized as a sedimentation basin. Good layout requires that the main gate in river-fed farms be located at the upstream of the main canal.

¹ Contribution to the FAO UNDP-SCSP Consultation Seminar on Coastal Fishpond Engineering, Surabaya, Indonesia, 4–12 August 1982.
² Aquaculture Engineering (Private Consultant), Iloilo City, Philippines.

Brackishwater farms in the region are so laid out that the main supply canal is also used as the drainage canal.

In one farm in Thailand, the width of the peripheral canal in relation to pond dimension is small. This does not provide better land use. The central pond platform used in the growing of natural food for the shrimps has not been excavated to maintain a certain depth of water at all times.

Water quality improvement would appear to be the most significant problem in most or all of the farms visited. There has been the need to treat the water for most farms before it is introduced into the culture system. The availability of the pretreated water in adequate quantity, however, remains to be the limiting factor in meaningful production especially where a hatchery and nursery units are present. The problems of water treatment by filtration should be provided where required and defective old filter designs should be corrected. Modifications of existing installations may make them workable and efficient.

There are three types of milkfish farm systems in the Philippines and these are the nursery pond system, the rearing pond system, and the combination of both nursery and rearing ponds. Milkfish pond layouts obviously have been designed according to the type of management practice, as layout and management are interrelated in fishpond operation. Principal systems of management commonly used are based principally on the type of food grown in the pond for the culture of the species.

2.2 Construction methods used in the region

Most fishponds owners do not construct the ponds completely. In areas where there is a limited capital, the construction proceeds as much as the finances will permit and only partial construction of the farm is done. Further, development of the remaining area depends on the income the owner makes from the fish harvests on the partially developed area.

Prior to actual construction, tidal observations are conducted, especially in the place where the main gate is to be constructed. The procedure involved in the determination and the establishment of the bench marks makes use of a plastic hose, about one cm in diameter and about 25 meters long. The hose is filled with water, taking care not to entrain air during filling. One end is held by a man at a station as marked while another man holds the other end at another station, which if along the dike centerline, is 20 meters away from the preceding station.

Dike construction begins by underbrushing, clearing and grubbing along the dike centerline previously established as staked and according to a prepared plan. Generally, the clearing is done along a five-meter strip for a main dike and along the centre of the strip, a trench about 50 cm wide by a meter deep or so, is dug and this is filled with mud taken from the borrow pit near and along the dike. Tamping of the clay puddle trench is done by foot or by the use of a heavy, square block of wood fitted with two handles.

The earth for the dike fill is excavated in rectangular blocks, each block about 50 cm × 30 cm × 30 cm, using a flat, steel blade about 60 cm long and about 15 cm wide at the tip and narrowing to about 5 cm at a T-handle. The implements used in the Philippines and in Indonesia are almost similar; in East Java, the digging implement is practically all steel, or all wood with a stubby metal blade. “Cangkul” or a modified version of a hoe but with a curved blade is used in Indonesia and that which is commonly used in Thailand is one with a long pole-like handle attached to steel blade that is flat as it is thin especially at the edges.

The soil cubes are neatly piled to form the dike; if there are roots within the block, the blocks in the centre of the dike are laid with the roots parallel to the dike line and blocks outside have their roots perpendicular to the line of the dike. A “receiver” hired by the owner, receives the blocks from the contractor's men, throws the blocks with force into the dike and stamps with his feet to seal the voids between blocks. The main dike is generally built in three stages, the first, lower stage to completely enclose the entire pond are in order to be able to impound water during a high tide. The presence of water in the impoundment allows the use of a “flatboat”, that facilitates the transfer of the earth blocks over a long distance while at the same time cuts down on time otherwise spent in walking through mud.

The main gate is constructed at the same time that the perimeter dike is constructed. The best time of the year for constructing the gates (and the dikes) in the Philippines is during the months of February and March when tidal waters are low.

Before the flatboat was developed by the Denila brothers in Iloilo, Philippines, the earth cubes were loaded on bamboo rafts. The shallow draft of the raft did not permit a bigger load per trip however. In Indonesia and other regions in the Philippines, dug-outs or small wooden boats are used for transporting excavated earth. The unloading process is rather slow for this type.

The flatboat has been considered as one of the best equipment for pond bottom leveling. When the lower ⅓ stage of the perimeter dikes has enclosed the area, water is let in to a level equal to the proposed elevation of the pond bottom plus 40 cm, and bamboo stakes are set at places all over the pond area as reference for height. A homebrew wooden depth gauge is used to determine the depth of excavation necessary. The earth blocks are loaded on the flatboat as they are dug-out, placed on the flatboat which is pushed towards the dike to make the second ⅔ stage or to make the fill for the secondary dikes. Unloading is easily accomplished by simply tipping the flatboat and this process saves time as there is no need to unload each individual cube.

The construction of the canals is done at the same time as when the adjacent dikes that make up the canals are constructed. Locating the positions of the stakes to mark the lines for slope and grade seems to be not too difficult for the experienced fishpond workers.

The use of mechanized equipment in fishpond development has been seen in Singapore and in Malaysia. Draglines and the power shovel have been proved to be the most suitable fishfarm equipment for earthwork in those countries (Fig. 1). Discussions on these equipments and their use are presented elsewhere in this paper.

Fig. 1 Range diagram and limits of work of drag line as defined for the construction of the perimeter dike

A wooden scow was once built in Iloilo, Philippines and this was used to build the perimeter dike of a fishpond being constructed on a reclaimed area at the shore. A backhoe which was fitted with a longer boom was installed on the scrow. When the depth of water during high tide was just right, the equipment was used to scoop out the bottom earth and dump it for fill on the dikes. The operation using the modified floating backhoe was so successful that the engineerdeveloper decided to build a ferro-cement scow to replaced the wooden one.

2.3 Physical assessment of existing farms

Tidal brackishwaters in some places have high silt content and this resulted to silt deposition on the pond bottoms at a high rate. The process of sedimentation has been so rapid that existing tambaks have gradually been reclaimed for urban development.

A project of Halong Bay in northern Vietnam has its main gate bottom higher than the elevation of the lowest low water. Complete drainage of the pond water was not possible. The same gate has a sluiceway width of 1.60 m and manual operation of the flashboards was very difficult (See Annex A and Fig. 2).

Fig. 2 Pressure diagram showing unit pressure (lbs. per sq. ft.¹) for a submerged vertical plane, the plane of the flashboards of a gate

Refer to Annex A for a sample problem on the computation of the total pressure on the bottom board (no. 8) and the total pressure on the sixth board from the top (no. 6 board) for a width of sluiceway equals to 4 feet.

¹ 1 lb/sq.ft = 4.9 kg./m²

The Halong Bay project could claim to have one of the best constructed dikes in all Asia. The dikes which have earth cores in them were rip-rapped with carefully laid stones. The two-door gates (2 units) have been found to be inadequate to tide over flash flood runoff from the nearby mountains during the heavy rains. Overtopping of the rear peripheral dikes at lakeside has been experienced and there has been a heavy disposition of silt at the sluiceway due to dike failure.

There are no supply and drainage channels that lead to and from secondary pond gates. Pond water exchange is hardly accomplished during the high and low waters. Absence of these channels resulted to poor management.

The concrete secondary gates were well-built except that they do not have the required number of grooves for screens and for flashboards. Complete sealing of the gate is not ordinarily attained inasmuch as the slabs do not provide a watertight situation as is ordinarily the case.

The tides are generally semi-diurnal throughout the year except for a few isolated days during which mixed tides are experienced. The suitability of the Halong pond for a tidefed farm in relation to existing pond bottom elevations is excellent.

Similar conditions exist in the Haiphong fishfarm. There exists a need to modify the present layout of the earth ponds. The layout of the experimental/production ponds of the Haiphong Institute of Research and Production, Vietnam, shown in Fig. 3 has the main supply canal bordering the entire pond area. It has been experienced that complete evacuation of the canal and pond waters have never been done. This has mainly been due to a large canal cross-section with a relatively flat slope and a narrow width of the sluiceway of the main gate.

The approximate length of the canal is 1 300 meters with a water cross-section area of roughly 18 m² side shot by a level made during a profile survey for a proposed drainage canal revealed that at some places along the canal, the bed elevations were below main gate bottom elevation, this condition would not make complete drainage possible.

To solve the problem of drainage, a network of new drainage canals cutting along a line through the centre of the entire pond area was proposed. A bed slope of 0.0012 was laid out on profile and the canal cross-sections drawn. Reference is made to Annexes E, F and G and Tables 1, 2 and 3, which give the basic theory of open channel design used in fishfarms.

Excessive seepage through the main gates was observed in a fishfarm in Indonesia. Investigation revealed the absence of cut-off walls. The length of the gate floor was shorter than what should have been required by safety against percolation. The wingwalls at the pond side were at 90 degrees with the direction of flow. This sudden change in the cross-section of the channel resulted in the eddying of the flowing water. The turbulence caused erosion of the sides of the earth dikes and of the earth canal bed.

Fig. 3 Map showing plan of existing ponds and proposed modifications for Institute of Shrimp Culture & Research Haiphong, Vietnam

Table 1
K values for trapezoidal channels in the Manning's formula (after Hechanova, 1977)

Table 2
Values of bed width, b for different values of depth D and side slope based on the formula (after Hechanova, 1977)

Table 3
Values of slope for different velocities V and depth, D (after Hechanova, 1977)

Channelization or the straightening of a stream was observed in a Philippine project. Investigations revealed the purpose which was to minimize local flooding by shortening the distance travelled and thereby moving the flood waters in the shortest time downstream. The benefit must be weighed against erosion and flooding of the area downstream.

It has been said that to excavate to a depth of more than 0.20 m it would not be economically feasible even if credit terms were liberalized¹. Some ponds in North Sumatra therefore, were sited at areas of high elevation and so can only be partly flooded during high water.

Potential acidic conditions of the subsoil down to a given horizon increase proportionately with depth. Potential acid sulfate soils are basically the soils of marine flood plains and elsewhere and along the coastal plains of Malaysia where “gelam” is widespread. Gelang Patah farm in Johore is one such project with soil acidity.

Pond compartments of some fishfarms have not been proportioned for relative sizes. Changes and modifications have to be made on some pond layouts as management methods change with time.

In a farm in Samut Sakhorn, Thailand, the width of the peripheral canal in relation to pond side dimension is small. The central pond platform which serves to grow the natural food for the shrimps does not contain water at all times as it should. Observations and study of the tidal conditions near the site is necessary to make the most of tidal flow in future design of the ponds.

There were some substandard installations of the physical plant observed in stations visited and not a single station keeps a log of operations, maintenance and inspection activities of equipment and facilities. The operators became lax and they forgot to do the necessary work as keeping a log book of operations and maintenance.

¹ PELITA III, FY 1979–84, North Sumatra Fisheries Development Programme.

3. SUGGESTIONS FOR IMPROVING THE ENGINEERING OF COASTAL FISHPONDS IN THE REGION

3.1 On design and layout

Brackishwater ponds are properly designed and laid out if all the pond compartments, canals and the water control structures mutually complement each other and thereby resulting in an optimum efficiency particularly in water management.

There is a need to modify the layout, if this is feasible, of some of the fishponds visited in view of new, improved methods and techniques of culture and, in some cases, to improve the structural stability of the farm components.

Proposals were made to rehabilitate the Babalan demonstration ponds in North Sumatra. The perimeter dike west of RP5 (Fig. 4) constricted the flow of high tidal water coming from the north and this situation resulted to an increase in flood stage at the vicinity of the constricted water-way. The portion of the main dike west of the nurseries failed by overtopping and by the subsequent erosion of the top of the side slopes. The surface velocity of flow at high tide was measured by the float method to be 3.87 ft/sec (1.18 m/sec), a value in excess of the minimum erosive velocity of 1.54 ft/ sec (0.47 m/sec) allowed for clay loam.

Fig. 4 Pond layout, Babalon Demonstration Ponds, Babalon, Langkat District, North Sumatra

Good design requires that the side of the dike facing the stream shall have a side slope of 2:1 or flatter and that the base be stabilized against rotational slips and/or erosion by providing some toe protection by means of continuous line bamboo culms or timber poles driven into the soil at the dike base. This however, was only a secondary solution to the problem. The jutted portion of RPI perimeter dike should be removed to provide a straight alignment of the waterway and a new dike constructed as indicated.

Further, inquiries were made as to why the constriction and the reason was probably to have a shorter span for the bridge, that which provides the access to the ponds from the centre buildings on top of the hill.

It has always been stressed that the most suitable place for the nurseries is where it can easily be supplied with new, unpolluted water all the time. The same figure (Fig. 4) shows the nurseries as indicated. Surely, when the rains are heavy, the runoff from the adjacent hill will wash the top side hill soil and this and other dirt will be carried to the farm into the nurseries. It was suggested that new nursery ponds be built at the east end of the farm where they can be easily supplied with new water every tidal cycle. The opposition presented to this suggestion was that the location would be far from centre management, the very reason in the first place why the nurseries were located at the west end.

The tidal inflow coming from the north into the Babalan farm in Indonesia was observed to have an appreciable velocity of flow due to constriction by a concrete highway bridge waterway. In general, dikes fail at vulnerable points like corners of the pond exposed to high flow velocities. The ability of the stream flow to detach the side slope material of the dike could be decreased by smoothing off the dike corners. avoiding sharp eroding bends in watercourses. Rearing pond, RP5 for instance, can be modified by removing the sharp corner and constructing a smooth bend. The same treatment should be done on the northeast corner of the farm.

The case of Prigi centre in East Java is ideal for the use of an SWS filter (Cansdale, SCS/79/WP/80), because its use removed the need for sedimentation and filtration of pumped raw water. A sedimentation and filter basin therefore, is not necessarily a requirement for the Prigi complex.

Pumping tests should be conducted to determine yield draw-down, recharge and quality of ground water. If results are satisfactory, a multi-stage pump of the required rating would be selected. A reservoir may be required to store a reserve flow when pump repairs are made. No safe estimates can be made on how the present well can tide over earth pond requirements for freshwater. There is the possibility that saline water may move into the freshwater aquifer and when this happens, termination of pumping is resorted to.

Shrimp ponds should preferably, have separate inlet and discharge sluice gates, drainage and supply canals. The sluice gates, if possible should be at opposite sides of the shorter pond dimensions to facilitate better exchange and circulation of pond water. To minimize pollution and spread of disease particularly when the ponds in a series are managed differently, there should be a separate canal for intake and for discharge of pond water.

Pond bottoms are sloped toward the drain, toward the discharge gate for a two-gated pond. The drain is laid out with ditches in a pattern similar to the veins of a leaf, the main ditch running toward the catching pond, if this is provided, or toward an area which is dug out deeper and of adequate area to hold the fish during harvest. If the pond bottom is well constructed, the pond drains easily without the ditches.

In large ponds with high perimeter dikes like the Satul farm, it is a good design if the pond side of the dike is provided with a berm, or a narrow platform. This facilitates inspection of the dike and of the cultured species as it would also break the erosive effect of the waves as they hit the dike. The berm is made deep enough to prevent predatory birds from alighting to prey on the cultured species.

For shrimp culture, it is not necessary to clear the inside of the pond from tree stumps; if the stumps are not to be removed, they should be cut short and are well below the level of the pond water. Milkfish ponds, however, will require that the pond be totally cleared to facilitate harvesting by gillnet seine.

For grow-out earth ponds which are built on less pervious soil, as in the case of the Chachoengsao ponds, means shall be provided to retain water or to reduce seepage losses.

Jamandre and Rabanal (1975), who had occasion to visit the construction site, mentioned in their report:

• that there was a need for a puddle trench and a proper and adequate compaction of the dike material

• that the dike side slopes were too steep, considering the type of soil in the site.

• that the base of the dike seemed to be narrow to be able to prevent lateral seepage

• that the paths of the dikes were not being thoroughly cleared of grass and other organic matter.

Ponds built on permeable, coarse-grained soils lack the proper amount of clay to form a seal. A 0.30 m thick clay blanket material containing at least 20 percent clay should be provided to cover the bottom of the pond. Before the blanket material is laid, all rocks, tree roots and vegetation are removed. Compaction using a crawler tractor or other suitable means is done on layers 0.15 m thick at “optimum moisture content”¹. This is necessary to attain maximum density.

A trench can be cut deep enough around the perimeter of the pond at the base of the dike and then is sealed with the clayey material. The pond side slopes are made flatter, about 3:1 if a film of polyethylene material is laid and covered with soil. The edges of the liner extend above the water line and the lower edges are anchored in the perimeter trench. Pond can often be sealed with time by manure or cowdung spread evenly over the pond bottom.

¹ From testing using the Proctor Model, it is the correct amount of moisture content that should be present in the soils during compaction in order to attain maximum density (and low permeability) after compaction.

Dikes piled up by draglines or pushed up by bulldozers usually do not get the required amount of compaction. Of course, good dikes can be built with these equipment if only attention is given to the need of proper compaction. Permeability is low at maximum soil density. Layout, especially of the perimeter dikes of Bedagai and Babalan farms in North Sumatra were observed to be defective and should be modified.

Main dike locations should be such that the following criteria are strictly observed:

1. The dike should be located on soils that provide the best foundation conditions and as a safety against seepage, shall be at a distance from a river bank based on the length of the line of creep¹. It is also important to determine the discharge, but if flood stages of the stream are not available, computations are made based on the hydrologic conditions of the watershed, rainfall conditions and frequency. The estimate of runoff can be made from these data. Philippine laws regulate the distance of the main dike from a river to be not less than 20 meters.

2. Avoid, if possible, creek crossing. A problem of excessive settlement and seepage will arise if a dike crosses an old stream.

Fig. 5 Soil grain size distribution curves
Soil A-Well groded having all sizes of grains in such a proportion that makes it easy to compact to maximum density
Soil B-Coarse fraction predominant
Soil C-Fine fraction predominant

Uniformity coefficient

3. Avoid reducing the width of the natural floodway as constricting the channel would adversely affect the safety of the dike and would cause increased flood stages on areas which are not protected against flooding.

4. Soil borings along dike centerlines are necessary especially for main dikes. The borings should reveal the depth and thickness of a permeable layer if there is any, to determine their influence on piping. For example, soils with high organic content will shrink due to oxidation of the organic matter and for these organic silts and clays, the preferred slope would be the angle of repose of the soil when wet. Inorganic silts have slow permeability and dikes made up mostly of this kind of soil should have flat slopes of 2:1 on the wet side. Organic clays have very slow permeability and dikes made of this kind of soil should also have flat slopes. Figs. 5 and 6 gave details on the classification of various soil types.

¹ See Section for of this report.

Fig. 6 Textural classification chart for soils (Based on US Bureau of Public Roads standards)

While a wooden sluice gate is ideal for grow-out or bigger ponds as it is practical and efficient, its use in small nursery ponds is not practical. The use of a pipe, wood, bamboo asbestos or ferro-cement for water supply makes it economical as a sluiceway for control of water supply and drainage. Culverts are usually underground pipe structures buried through the dike and are designed to handle quantity requirements for supply and drainage. The conduit itself should be free from leakage due to internal pressure. Anti-seep collars should be used as an added protection against seepage along the surface of the conduits. The discussion on design of culverts used in fishfarms are presented in Annex B, C and D, Table 4 and Fig. 7. Circular, concrete pipes were considered in the discussion.

The floor of the main gate should be as low as possible, lower than the lowest low water experienced in the area. This is to ensure complete drainage, provide water cushion for the pump and for the tidal outflow, and to eliminate the possibility of undercutting below the gate bottom.

Fig. 7 Details of culvert installation.

Table 4
Capacity of concrete pipes with submerged outlet, culvert flowing full in cu. m/sec

Values are based on the formula:

Q = 2.093 D²H^0.5 for concrete pipes in which
Q = discharge in cu. meters per second
D = diameter of pipe in meters
H = head on pipe, meters = difference in elevation of water surface at inlet and outlet ends

Capacities do not include friction loss because of short pipe lengths observed in practice. The slope has no effect on the capacity, the “head” being the controlling factor.

Most gates have not been proportioned to the area it serves to flood and drain. The determination of the width of the sluiceway seems to be a common problem for some developers, although for people who have the experience, the problem has been looked into with less difficulty. Since flooding of the ponds is done during high tides, the determination of the width of the sluiceway should therefore be based on tidal range, canal and pond areas, and the desired number of successive high tides that the ponds be flooded. An analytical method of the determination of the width of the sluiceway of the main gate was presented by Kato in 1975 and a sample of the design used for a Malaysian project has been described (Hechanova and Tiensongrusmee, 1980).

The width of the sluiceway for all main gates should be of the same dimension. This is to allow interchanging of the flashboards. The specifications for appropriate flashboard are given in Annex A and H and Fig. 2.

Excessive seepage through the main gates in the North Sumatran project was due to the absence of foundation cut-off walls and the inadequate length of floor. For conditions where a gate is built on pervious material, the floor may be given a minimum length, a minimum dimension of twice the maximum full head of water measured from the floor level to the level of the highest tide. The path of percolation is formed by the cutoff walls and the underside of the floor, the upstream cutoff wall if on pervious foundation is usually carried to a depth below the floor to about equal to the height of the water surface above the floor and the depth of the down-stream cutoff wall usually extends to a depth about equal to the depth of the water in the canal. For impervious conditions, dimensions of depth may be taken as one-half of the depths required for pervious cases.

Hydrostatic uplift pressure should be considered and stability against uplift is to be provided.

Where masonry or bricks are available locally, the side and wingwalls are, for economy, made of these materials. The brick side wall is designed as a gravity wall, the soil face of the wall is stepped and the weight of the wall and of the earth on top of the steps prevent overturning due to lateral earth thrust. Gates of brick masonry are suitable and economical for farms in Malaysia and Indonesia where bricks are locally produced in quantity.

Gate wingwalls provide the transition from the concrete sluiceway to the earth canal. Since the earth canal velocity of flow must be lesser than the flow through a concrete sluice-way, the earth canal bed is made wider than the width of the sluiceway. These wingwalls allow enlargement of the canal cross-section necessary for a reduction of canal flow velocity. To effect a reduction in the velocity of flow as water reaches the earth canal, the transition of the pond side of the gate provides the gradual enlargement of the channel cross-section. Sudden change in the cross-section results in turbulence and the consequential erosion of the dike side slope and the canal bed. Gate wingwalls at the river side of the gate may not necessarily be at an acute angle with the direction of flow as a 90-degree bend is allowable.

The overpour type is best adopted to main gate regulators where water from the source carry large amount of sediment. With this type, surface water in the pond could be drawnoff during periods of heavy rains. The use of the undershot type gate has one special advantage as in this the sedimented material in the canal can easily be washed out during drainage and that bottom water can be drained out when pond DO levels are low.

A pump which may be integrated into the design of the main gate will require that the gate must have at least two openings, a pump sump and provisions for the screening of the raw water during pumping. The sump is usually in one of the sluiceways and the water is discharged into the other sluice-way. Sump bottom elevation must be such that there should be adequate submergence of the intake pipe end during low tides. The other sluiceway into which the water is discharged must have water at all times at a certain depth necessary to provide cushion to the falling water. This depth is generally equal to 1/3 of the height of fall.

Wood has been a widely-used material for secondary and tertiary gates and well-designed gates of wood were observed in most places visited. Modifications however are suggested with emphasis on some main points.

• The cleats to which the side boards are nailed should be on the sluiceway side to prevent future removal by lateral earth pressure. The holding power of nails decreases with time and so can not be relied upon if the side boards are nailed on the inside.

• Anti-seep boards at both sides are essential and cutoff walls are provided at both ends of the gate to prevent water seepage and undermining of the gate.

Complete sealing of the closure boards is a frequent problem and one common procedure is to pack earth between two sets of flashboards. These two sets of flashboards are necessary in all gates as the boards can also be adjusted to let surface water drain over the first set and bottom water drained off the other set by removing the bottom board of that set.

The safety of the main gates in the case of Satul fishfarm against washing at the sides depends to some extent on the care taken in backfilling of the sides. Concrete anti-seep collars should be added to the side walls of the gates of the Satul farm as a permanent solution to seepage and consequent wasing of the backfill.

The location of a river intake structure is a principal consideration in the layout of a fishfarm canal system. A good layout locates the intake at a straight river stretch where erosion and turbulence are at a minimum and as near to a fresh tidal water supply. River intake structures are never located at a stream oxbow, where at this point considerable filling up occurs at the inner bank and erosion occurs at the outer bank. An intake located at the delta section of the stream could present the problem of heavy silt deposition.

A sluiceway stretch should be provided between the river intake and the main gate. This has not been the case in farms visited. This stretch is called the idle or dead reach and is designed to collect and remove from the water the silt which is liable to be deposited on the canal bed. The dead reach consists of an enlarged basin in which the velocity of water is decreased to cause deposition of the transported material, the amount of which will depend on the reduced velocity and the length of time the water is in the basin. This length must considerably be at least six times the depth of water in the canal.

A knowledge of the character of the stream flow and of the silt is desirable. Difficulties are encountered in farm operation and management due to a heavy load of sediment in water. The economic life of the farm and for many of the structures has been much shorter than anticipated. The silting up of the canals and of the ponds make them inefficient and very costly to maintain. Minimizing silt loads would be the basis for proper design rather than desilting. Siltation of the main water supply channel and ponds of the Bangil farms in Kalianyar in East Java and the surrounding pond areas is one of the problems. The silt load of Porong River which traverses the area was observed to be apparently heavy as indicated by its very high turbidity and sluggish flow. No information was available on how severe is the sediment problem but definitely the situation is disturbing. Silting up of the river bed would increase flood stages and would create more flooded areas. A knowledge of the character of the stream flow and of the silt load is desirale as, by careful planning, the disturbing results may be avoided or minimized. Samples are taken in rivers under all conditions of river flows and the total sediment load of the river calculated.

One principal consideration in layout should be such that the least amount of silt is carried into the canal. To obtain this result, partly, the canal is made perpendicular to the river centre line and the main gate located further upstream to the ponds. The sluice channel forms a basin in which the water moves at a decreased velocity due to a higher positive canal bed slope and that the flood of the sluiceway channel kept as low as possible, if feasible with respect to the gate bottom. The bed slope is such as to develop a scouring velocity during downflow of not less than 5 to 10 ft/sec (1.5–3.0 m/ sec). The floor of the sluiceway channel must therefore be lined and the canal sides provided with grouted riprap. Refer to Annexes E, F and G for discussion on canal designs that give the required scouring velocity during low tidal flow and settling velocity during high tides.

Dikes constructed on very soft organic silt present a problem of cost and stability of the base. It may be economical to excavate the soft material if the thickness of the soft stratum does not exceed 1.5 m, otherwise, if the thickness is greater, it is preferable to permit the sinking dike to displace the soft material which process could take many years. To shorten the period of settlement, the fill may be built up to an excess height above the final grade.

In order to reduce seepage flow in the North Sumatra project, an attempt by the North Sumatra Fisheries Service was made by placing vertically, a continuous sheet of plastic material to a depth of 2 meters (width of plastic sheet), a meter distance from the dike toe. When a trench was dug, piping was observed below the two-meter depth. This obstruction by the plastic cut-off wall was not effective as piping occured below the two-meter depth. This cutoff must have caused a concentration of flow lines and these lines must have possibly magnified subsurface erosion.

One consideration in layout of a perimeter dike is to locate the dike centreline as far away from the river bank, this distance determined on the line-of-creep approach to the problem. Values of creep ratio, C have been determined for most principal types of soil and with known values of head difference between the pond water surface and the river surface at low tide, the line of creep, Lp is computed. This line of creep, L approach to design has gradually been recognized to contribute toward reducing the danger of piping. On this account therefore, it would be on the side of safety if the peripheral dikes are located at least equal to this length of the line of creep (Fig. 8).

Fig. 8 Diagram indicating dimensions for computing length of line of creep (with reference to North Sumatra project discussed in text)

In Songkhla farm for instance where the tidal range is low, the width of the sluiceway should be made bigger to allow the passage of the maximum tidal waters during flood periods. The pumps supplement the water quantity required to fill the ponds to the desired water level.

¹ In Fig. 5, the length L = t₁ + B, and this distance L must be as great as C_c × h. See Piping, Case Study (Section 4), for more discussions on creep.

Channelization or the straightening of a stream has been thought of as a means to minimize local flooding by shortening the distance travelled and thereby moving the flood-waters in the shortest time downstream. Jamandre and Rabanal (1975) gave engineering recommendations for the provision for drainage of flood waters and it was suggested that straightening the water courses can be helpful.

Channelizing a stream could result to an increase in erosion due to abrupt increase in the gradient. The widening and deepening of the channel could cause serious problems along the banks. Although channelization has enable more flood plain land to be utilized, this benefit must have to be weighed against erosional loss of the adjacent farmlands and the farmlands downstream due to extensive flooding.

Marine research preserves should be established to protect the mangrove and to serve as ecological baselines by which man-induced changes may be evaluated.

The cost of a project is influenced by the requirements of the design and of the specifications. For instance, secondary gates must be all of the same design and specifications, and so with tertiary gates. The use of local materials should be preferred over other materials which are not economically available. Where brick or masonry are plenty, main gates could be of masonry/brick construction without necessarily resorting to a totally reinforced concrete structure. Structure design should be simplified where possible so that they are relatively easy to build.

3.2 On methods of construction

The advantages of mechanized fishpond construction have been found to be limited and the cost is relatively high (Denila, 1977). A low ground pressure swamp dozer which was used in the construction of the SEAFDEC ponds in Leganes in the Philippines had bogged down everytime it was on soft ground. An LGP dozer brought into the Philippines for use in a project in Lanao (FAO/UNDP/BFAR) in southern Philippines was never used as there was the difficulty of moving it to the site. The cost of hiring machinery for earthwork has been one of the major constraints in fishpond construction.

Singapore and Malaysia generally use dragline to excavate earth. One of the reasons for its use is the fact that there is a lack in availability of manual labour with skills in pond construction. The particular equipment is particularly good for digging of canals and building the dike as it does not have to go into the pit in order to excavate. It operates on natural ground while excavating materials from a canal and is more suited especially when in wet soil. The excavated earth can be deposited along the canal for the building of the perimeter dike.

A crawler mounted dragline was proposed for use in the Ban Merbok project in Malaysia, to excavate earth from borrow pits along the dike and deposit the material to build the dike. The illustration in Fig. 1 shows the limits of work and the range diagram as defined for the construction of the perimeter dike. As work progresses, a mat of logs from trees cut in the area forms a stable base and prevents the equipment from bogging down in mud. When draglines are used for building the dikes, the disturbed material are simply permitted to establish their angle of repose and therefore the side slopes are made more flat at about 2:1 to 3:1

Malaysian operators were observed to be very skilled in operating a dragline. Excavating is started by swinging the empty bucket to the digging position, pulling the bucket toward the machine while regulating the digging depth. When the bucket is filled, the operator takes on the hoist line while playing out the drag cable. Hoisting, swinging and dumping of the loaded bucket follow in that order; then the cycle is repeated. Fig. 9 shows how the type of soil material and the size of the bucket could affect the cost of excavating earth using a dragline. A dragline can make two cycles per minute, and a 2-cubic yard bucket can handle 230 cubic yard per hour, (175 m³/hr), bank measure. For wet sticky clay, the output could be about 145 cubic yard (110 m³) for a short boom dragline. The cycle time however is reduced for 70-foot booms (21.4 m) with a 2 cubic yard (1.53 m³) bucket.

Fig. 9 The effect of the class of material and the size of the bucket on the cost of excavating earth using a dragline (Source: Peurifoy, RL., Construction Planning, Equipment & Methods, 1970 McGrow, ISE, Tokyo) (1 cu. yd.= 0.76 m³)

It is important that before selecting the equipment to be used, the job must first be analyzed as haphazard selection of equipment can result in increase in cost of earthwork.

Dozer-crawlers with LGP tracks come in variety of capacities from 20 BHp to as large as 200 BHp. For sites that may require difficult transport of equipment and because of limited lane loads of rural timber bridges, a small LGP may be chosen. A dozer operator should know the economical range of his equipment. For instance, a crawler is most effective pushing earth in short distances and uneconomical if the distance is beyond 50 meters. Beyond this distance, if the volume of soil hauled is big, payloaders may be employed instead.

4. PIPING: A CASE STUDY

A typical failure caused by piping was observed in a pilot project in North Sumatra which resulted in the total loss of water in the newly built ponds. The water that escaped from the ponds through the subsoil came out of the ground in the form of small springs downstream near the bank of the river.

Piping is the sudden formation of a pipe-shaped discharge channel or tunnel located between the dike foundation, the overlying impermeable scattered layer of clayey material and the permeable soil.

Piping in the particular project was seen to be due to scour or subsurface erosion that started at the springs near the toe of the dike. The mechanics of piping is such that the sub-surface erosion progresses and proceeds upstream into the pond along a bedding plane of the overlying clayey material. Failure due to loss of water from the impoundment occurs as the intake end of the eroded hole reaches the bottom of the pond. Occurence of excessive piping was evidenced by the formation of springs at downstream of the dike where several outcroppings were seen. A sample of the cloudy water that came out through one of the tunnels was taken and this was let to stand for a couple of days. It was observed that the water cleared when sedimentation took place completely. The sedimentary material was investigated and was found to be non-cohesive silt with fine sand fractions. The progressive removal of soil material through the tunnels and through the springs was apparently a case of piping.

It was previously assumed that the leaks were attributed to a physical and chemical quality of clay, termed “dispersive clay”, inherent in fine-grained soil which are highly erodible. These clay types are not easily recognized by field identification tests currently employed in civil engineering practice. However, laboratory criteria for identifying these clay types have already been developed.¹

A substantial settlement was observed of the main dike and this was evidenced by a vertical displacement of 0.50 m below the floor of a wooden bridge which was at the same level with the dike top originally.

The main dike was constructed very close to the river and some stretches of the dike crossed an old creek. The foundation material below the main dike was a deposit of noncohesive, organic silt with fine sand fractions².

Dike foundations at most places along the main and secondary canals were reported by the workers to contain a number of tree trunks and stumps which may have been left during clearing activities. Facilities for removal of these stumps may not have been available during clearing and grubbing.

Burrowing crabs and animals and their holes were seen along the perimeter dikes and at the pond perimeter canals. Earth mounds of Thalassina were seen scattered along the dike sides, on the tops and in the ponds. The holes were about 0.05 m in diameter and when probed, were one meter deep and at an incline into the subsoil. The holes were also observed to be the percolation points.

4.1 Means taken by the North Sumatra Fisheries Service to minimize seepage

In order to reduce or control seepage flow, an attempt was made by placing a continuous vertical wall of plastic sheet to a depth of two meters (the width of the plastic material) at a distance of one meter from the dike toe outside the pond.

At the dike base inside the pond where there was sandy material, reinforcement of the base was done by dumping soil taken from the pond peripheral canal.

The dike base was probed for tree stumps, boulders, etc., and these materials when found, were removed and the gap replaced with soil material from the pond.

4.2 Suggestions and recommendations

Technical assistance was given to the North Sumatra Fisheries Service to correct and offer solutions especially to the brackishwater engineering problems of the Bedagai fishfarm, particularly that of seepage.

A high and massive new line of fill taken from borrow areas at the pond central platform may be dumped into the thick layer of silt deposit at the old creek bed next to where the old dike had crossed. To shorten the period of settlement, the fill may be built up to an excess height above the final grade. This was however, seen to be impractical as no borrow area with suitable and adequate material for the fill was seen anywhere, except at the pond central platform.

It may be more economical to remove the thick layer of silt before placing the new fill if the thickness of the soft stratum does not exceed about a meter and a half. Removal is to be effected by the use of a mud pump working simultaneously with a strong water jet. The watery silt deposit is allowed to flow downstream, through a break made in the old dike.

Clay cutoff trenches may be constructed along the upstream (pond side) base of the main dike. These are trenches with sloping sides, extending through the previous layer into the less previous stratum, if this is feasible. An upstream blanket of clayey material is laid over the floor of the pond. The underlying layer is locally removed or excavated to permit construction according to the cross-section shown in Fig. 8. Borrow material for the clay blanket would be excavated from the central pond platform where clayey material is available. Soil samples were analyzed and the test results showed that the material was suitable. The quantity however may not be enough.

¹ Pinhole Tests for Identifying Dispersive Soils, Journal of Geotechnical Engineering Division. January 1976, GTI.
² The soil auger was extended to two-meter in length and because the soil material had less tendency to cling to the auger while wet, an open trench was dug to more than two-meters of depth inorder to obtain a sample.

Grout curtains may be resorted to if clay-filled trenches could not be extended deep into the impervious material. Grouted cut-offs are produced by filling the voids of the subsoil with cement, clay, chemicals or a combination of these materials by injecting the substance into the subsoil under pressure. Grout curtains have some shortcomings though, as neat cement does not penetrate the voids of a granular material unless the effective size of D₁₀¹ exceeds 0.50 mm if loose, to 1.40 mm if dense. Some portions of the subsoil may escape penetration by the grout material. Grouting operations depend on the skill and experience of the grouting personnel. Grout mixtures of clay and water, or clay, cement and water may be used to provide the necessary barrier to flow. When a deposit of clay is used as a grouting material, it could prove to be economical as it can be mixed with water to any desired consistency and injected using grouting equipment. Cement-clay grout is mixed in the proportions of 1½ parts of cement, 7 parts of clay, and 6 parts of water by volume. The clay is screened prior to mixing.

Cut-off walls of plastic sheets may be used if the sheet (two meters wide) could extend through the permeable layer into the impermeable or less permeable stratum. The situation becomes critical however, if the sheet does not reach the impervious stratum as there would result a concentration of flow lines below the plastic sheet, thus magnify the degree of erosion due to increase in velocity of flow of water through the soil voids.

In cases where the springs or outcroppings are substantial in size, the exit of the soil particles are prevented by the use of a granular inverted filter. This granular filter of carefully graded sand; the exit of the soil particles are prevented by the filter which has granules fine enough to prevent the escape of the soil particles that flow with the seepage water but coarse enough to allow the water to flow through. Filter design requires patient experimentation though, as careful gradation of the filter is important. Economic considerations may not warrants its use.

Existence of dispersive clays have undoubtedly been recognized in the North Sumatra fishfarm. If laboratory tests² will show the soil to be dispersive, liming would give positive results in that the addition of 4 percent (by dry weight of soil) of hydrated lime would convert this dispersive-clay soil into a non-dispersive erosion-resistant soil. Specifications require that the dispersive soil to be treated shall be broken down thoroughly by discing or pulverizing with a rotary mixer prior to adding the lime. The soil is then brought to optimum moisture content³ mixed throughly with the lime, allowed to cure in the loose state for a minimum of four days and then compacted. A blanket of this treated soil is placed over areas where seepage occurs.

Perhaps a new and modified layout may be planned such that no dike would cross an old creek, and where the required length L, of the “creep” line is satisfied for dikes on previous foundation⁴. Using this line of creep approach⁵, in the case of silt with fine sand fractions the equation for determining the length of this line, L, (Fig. 8) is,

The required distance of the dike from the river using the dike from the river using the above equation is found to be 51 meters. A new line of peripheral dike 50 meters from the stream, was suggested to be constructed, and that the pond areas outside this new dike may be abandoned.

Extensive soil sampling should be undertaken along this chosen line and the vertical profile of the substrata recorded to show the depth of the permeable layer if there is any.

For preliminary soil survey, Table 5, Field Determination of Physical Characteristics of Soils, will assist in evaluating the soil conditions along the line of the proposed dike and can be used as a guide to soil stability and permeability. The reader is referred to the discussion on design and construction in this report.

5. SUMMARY AND RECOMMENDATIONS

An increase in engineering input is necessary to increase production per unit area. Information on most of the engineering technologies have already been recognized by the aquaculture industry but there is still the need to fill in some technology gaps, for instance.

1. There is a need to simplify the concepts and principles involved in the planning and design of pond structures, for the non-engineer fishfarmer to be able to apply these principles in practice.

A simplified field manual or handbook on Aquaculture Engineering for the South China Sea Coastal Fishfarms may be prepared for use of the aquaculturists.

Minimum design requirements will be listed for each structure design including specifications.

¹ The effective size, D₁₀, of a soil is the largest diameter of the grains comprising the finer 10 percent of the soil. The percentages of grains finer than any given diameter are plotted as ordinates to an arithmetical scale and the grain diameters on a logarithmic scale. See Fig. 5.
² Journal of the Geotechnical Engineering Division, “Pinole Tests for Identifying Dispersive Soils”, January 1976, GTI.
³ The optimum moisture content according to the Standard Proctor Test is the value of water content in the soil during compaction and after which, the dry density is maximum. Compaction at optimum moisture content results to a very dense, and less pervious soil.
⁴ Line of creep is the path that a water particle follows along the surface of contact between the overlying less pervious material and the pervious soil.
⁵ Modified for stratified conditions at site.

Table 5
Field determination of physical characteristics of soils

(Source U.S. Soil Conservation)

Survey techniques for cadastral, topographic, hydrographic, biological, water supply and soil surveys will be discussed further and the standard methods laid down.

A standard design of aquaculture structures should be made available with particular emphasis to what local materials and technical skills are available in the different areas of the region. Specific designs would be dependent upon what resources and upon what appropriate aquaculture techniques are available.

2. Uncontrolled fishpond development causes upstream flooding that affects urban and populated areas. The indiscriminate loss of mangrove areas could create ecological damage to fish sanctuaries.

Environmental impact studies should be required prior to development with a view to conserving areas not suitable for fishponds.

3. There is a need for more information and the dissemination of information on the problem of acid sulfate soils and how to overcome the problems met.

4. Based on the North Sumatra experience, soil quality considerations deserve priority attention particularly on the prediction of seepage which will require the proper approaches.

ANNEX A
HYDROSTATIC PRESSURE ON FLASHBOARDS OF A MAIN GATE

Example problem:

A flashboard type of gate control has eight 12-inch (30.5 cm) wide flashboards. The width of the sluiceway is 4 feet (1.2 m). It is required to find (a) the total hydrostatic load on the bottom board and (b) the total load on the 6th board from the top.

Solution: The pressure diagram is shown in Fig. 2, remembering that unit pressure varies directly with head “h” (in our case, depth of water). The pressure diagram takes the form of a triangle.

Then,

The total hydrostatic load, F = whA, on.

In Fig. 2, the shaded area represents the load pressure on flashboard No. 6. The total force is the average of the unit pressures acting on the bottom and on the top edges of board No. 6 multiplied by the area of board No. 6.

ANNEX B
DESIGN OF CULVERT FOR SIZE AND SLOPE

The design of culverts used in fishfarming will be limited to circular concrete pipe culverts in this discussion both for unsubmerged and submerged flow.

In engineering practice, the term culvert is commonly applied to any large underground pipe used in short lengths where a water channel crosses an embankment, such as a dike.

Hydraulic considerations.

1. Outlets are either unsubmerged or submerged.

2. Water surface at inlet may be at top of pipe or at a certain head above the pipe invert, where a monk with flash-boards is provided at one end.

3. Proper installation, alignment and grade are to be adhered to in order to give the service it is intended for its design.

4. Culverts flow full or partially full.

5. The coefficient of roughness, n, for concrete is 0.021.

6. Entrance and exit conditions are the same.

7. Whether or not the culvert runs full depends on the diameter, length, roughness and the level of head and tail waters.

8. The slope has no effect on capacity, the head being the controlling factor.

Design requirements:

1. The cross-section of the embankment or dike at the culvert site must be known.

2. Information as to rate of flow, m³/sec through the proposed culvert should be determined. Total drainage is to be completed during ebb flow of tide.

3. The maximum depth of water in the pond must be known.

4. Culvert must be large enough to carry design floods in a relatively short period of time.

ANNEX C
PIPE CULVERTS WITH UNSUBMERGED OUTFLOW

The flow of water at critical velocity for maximum discharge at any cross-section of a channel is that due to a head equal to half the average depth of the water at the cross-section. (Woodward, Technical Report No. 7, p.149 “Hydraulics of the Miami Flood Control Project”). Therefore, applying this principle to a circular pipe,

h = 0.3113 × D, where D is the pipe diameter, meters
      h is the difference in elevation of
      water surfaces, meters

Critical velocity therefore is,

With the area and velocity known, the discharge, Q = Av

A (in terms of 0.6887D, in figure) = 0.5768 × D²
Q = 0.5768 × D² × 2.47 × D½
= 1.424 × D^5/2

The slope on which a culvert pipe must be laid is computed using Manning's formula and, for a given diameter of pipe in inches, the slope is found from the following table.

ANNEX D
PIPE CULVERTS WITH SUBMERGED OUTLETS AND FLOWING FULL

Culverts are never designed to flow full but under fishfarm conditions, the outlet of the culvert, may be submerged, as when obstructed by tide flood water or water in the drainage canal. In this case, the culvert will flow full and the required size of the culvert can be determined from Table 4.

Table 4 is based on the formula for submerged flow, using a coefficient of discharge, C = 0.602 for submerged, sharp-edged circular orifices. The effect of length L of the culvert and a loss in head are neglected due to small values of L in relation to pipe diameter.

where Q = cu. m/sec
D = inside diameter of pipe, meters
h = head, meters, expressed as the difference in levels of the head and tail waters
A = cross-section area of pipe, m²

Table 4 gives the values of Q for different values of inside diameter, D of pipe and for head difference, h.

Numerical example:
Concrete pipe culvert with submerged outlet and flowing full. Head difference of pond water surface and of water surface in canal is 0.60 meter. From Table 4, if the inside diameter of the pipe is 0.30 m, the capacity Q equals 0.145 m³/sec.

To drain a 0.10 m of rainwater for example, during an excessively heavy rainfall from a ½ ha nursery pond, it will take,

ANNEX E
DESIGN OF OPEN CHANNEL CROSS-SECTIONS (UNIFORM FLOW)
AS APPLICABLE TO TRAPEZOIDAL EARTH CANALS IN FISHFARMS

Basic concepts

In practice the economics of canal design is affected by the following factors:

1. If the cross-sectional area, a is to be a minimum, the velocity v is to be maximum. Scouring of the erodible canal bed will occur when velocities are high.

2. Canal capacity can better be increased by widening than by deepening.

3. The minimum value of the canal area, a includes overburden. Therefore, minimum a does not imply minimum excavation.

4. The cost of excavation is not solely dependent on the amount of earth removed. Considerations as earth disposal and/or transfer may be important as the volume of earth material excavated.

5. Side slope could vary if design takes into account minimum seepage.

6. For a given canal section, the slope of the canal bed determines whether the velocity of flow is silting or transport velocity. It is desirable that velocities in earth canals be high enough to transport the silt particles during drainage without producing erosion of the bed and the side slopes. During flooding of the ponds, the flow is sluggish due to a canal upgrade. The silt laden tidal water in the main canal is therefore partly freed of the silt before water goes into the ponds. The removal of these silt particles that have settled on the canal bottom is necessary during drainage.

Permissible canal velocities for different soil types

(Source: Fortier and Sobey, “Permissible Canal Velocities”)

Procedure for design of a trapezoidal earth canal by use of tables

1. Determine the discharge Q in m³/sec. This is calculated depending upon the following:

   1. Water supply requirements

   2. Adequacy for drainage especially in times of flood or excessively heavy rainfall.

2. The desired depth of water in canal is determined accoring to:

   1. Depth of water in canal when at high tide
   2. Depth required if canal is used as shelter for the temporary holding of fish or for fish stocking

   3. Depth required if canal has the function of regulating the water within the farm if soil improvement by leaching is to be done.

3. The value of canal bed width,b as assumed is taken from Table 2.

4. Solve for the ratio of D to b. Table 1 gives “K” values for known D/b ratios and canal side slopes.

5. Determine the slope, from Table 3 for required velocity v and given depth d. Permissible velocity values are obtained from above Table (text).

6. No table for values of discharge Q in m³/sec for known values of K, S and d was prepared for lack of time. Coefficient of roughness, n equals 0.0025. The value Q is computed using the formula, Q = K/n × d^8/3 × S^1/2

Numerical example

Given the following data:

Discharge, Q = 6.00 m³/sec
Desired depth, d of water in the canal is 1.00 m
Canal side slope is 1.5/1

Solution:

1. From Table 2 find the value, b for maximum hydraulic efficiency, b = 0.604 m for a canal water depth of 1.00 m and a side slope z of 1.5/1. Use b = 3.00 as required for fish holding.

2. Solve for the ratio d/b. d = 1.00 m and b = 3.00 m 1.00/3.00 = 0.333 Table 1 gives K = 3.512 for d/b = 0.333 and side slope, Z = 1.5/1.0

3. For clay and for a depth, d of 1.00 m, canal bed slope Z is 0.00197, from Table 2.

4. Compute the value of Q, using the formula
   Q = K/n × d^8/3 × S^1/2
       = 3.512/0.0225 × 1.00^8/3 × 0.00197^1/2
       = 156 × 1.00 × 0.0443
       = 6.91 m³/sec. This is greater than the required discharge Q of 6.00 m³ sec but is acceptable.

ANNEX F
OPEN CHANNEL (TRAPEZOIDAL EARTH CANAL) DESIGN

Example problem

It is required to determine the required depth,d and the velocity of flow v of water in a trapezoidal earth canal of width b equals 3.00 meters, n equals 0.02, bed slope s equals 0.001 and canal side slope, Z of 1:1 (horizontal to vertical). Discharge, Q is 6.00 m³/sec.

Solution:

This problem is solved by trial. Assume a value for d and compute the values of a, p and r. Find v by Manning's formula and compute the discharge Q for each trial. Refer to sketch below (Note: Manning's formula is an internationally accepted formula for the expression of uniform velocity of flow, v in open channels, expressed as v = 1/n × r^2/3 × s^1/2, where n is a roughness coefficient of the channel, a is the water cross-section area, p is the wetted perimeter defined as the length of the line of intersection of a cross-sectional plane with the wetted surface, and s is the canal bed slope expressed as the ratio of the vertical rise to horizontal run. The hydraulic radius is the expression, r = a/p.

Trial 1

For d1 =	0.50 m
a1 =
P1 =	3.00 + 2 (0.50 × 1.414) = 4.41 m
r1 =	81/p1 = 1.88/4.41 = 0.426 m
v1 =	1/0.02 × 0.4262/3 × 0.0011/2
=	50 × 0.566 × 0.0316 = 0.89 m/sec
Q1 =	v1a1 = 0.894 × 1.88 = 1.68 m3/sec

Trial 2

For d2 =	1.00 m
a2 =
p2 =	3.00 + 2(1.00 × 1.414) = 5.83 m
r2a2/P2 =	4.00 × 5.83 = 0.686 m
v2 =	1/0.02 × 0.6862/3 × 0.0011/2
=	50 × 0.777 × 0.0316 = 1.22 m/sec
Q2 =	v2 × a2 = 1.22 × 4.00 = 4.88 m3/sec

Trial 3

For d3 =	1.50 m
a3 =
p3 =	3.00 + 2(1.50 × 1.414) = 7.24 m
r3 =	a3/p3 = 6.75/7.24 = 0.932 m
v3 =	1/0.02 × 0.9322/3 × 0.0011/2
=	50 × 0.954 × 0.0316 = 1.51 m/sec

Q₃v₃ × a₃ = 1.51 × 6.75 = 10.19 m³/sec

Plot d against Q for trials 1, 2 and 3 and read the value d equal to 1.15 m for the given Q of 6.00 m³/sec. See sketch.

Allowable maximum transport velocity without erosion of the canal bed and sides, for alluvial silts with colloidal particles is 1.52 m/sec. (Fortier and Sobey, 1926, Permissible Canal Velocities, Transaction of ASCE 89:955). At canal desired water depth of up to 1.50 m and less the canal velocity of flow is 1.51 m/sec and less, this being less than the maximum allowable of 1.52 m/sec, the design is acceptable.

In most cases, the main supply canal is used simultaneously for pond water drainage although a better layout is to have a separate drainage canal. Main canals are also used for holding and sheltering fish, and the size for required width and depth of water depends on the quantity of fish to be held. Generally, one cubic meter of water is adequate for every 1.5 kg of fish held in the canal during harvesting.

The main canal must also be designed to handle excess pond water during an excessively heavy rainfall. For example, if the maximum recorded daily rainfall distribution is 0.20 m, a 30 ha pond is drained of this excess water Q during the time of low water in about 4 hours or 14 400 seconds.

The main gate is also designed to adequately handle this excess water held in the ponds during a heavy rainfall.

ANNEX G
OPEN CHANNEL (TRAPEZOIDAL EARTH CANAL) DESIGN

Example problem

Given a trapezoidal canal cross-section, shown in the figure, n = 0.02 side slope, z = 2/1 = 2 and bed slope, s = 0.002. It is required to determine the discharge, Q in m³/sec, and the velocity of flow, v in m/sec.

Solution:

1. Find the cross-section, a: side slope, z = 2/1 = 2
   a = bd +(z) × d²
      = 4.00 × 1.00 + 2 × 1.00² = 6.00 m²

2. Solve for the wetted perimeter,p

   

3. Solve for the hydraulic radius, r r = a/p = 6.00/8.47 = 0.708 m

4. Solve for velocity, v
   v = 1/n × r^2/3 × s^1/2
      = 1/0.02 × 0.708^2/3 × 0.002^1/2
      = 50 × 0.794 × 0.0447
      = 1.77 m/sec

5. Solve for discharge Q
   Q = v × a
       = 1.77 × 6.00 = 10.62 m³/sec

Canal side slope, z is the ratio of the horizontal to the vertical. The bed slope, s is the ratio of vertical rise to horizontal run. A bed slope of 0.002 has 2 meters rise in a run of 1 000 meters (2/1 000).

ANNEX H
DETERMINATION OF THICKNESS OF FLASHBOARDS OF A MAIN GATE

Example

It is required to determine the thickness t, inches of the wooden flashboards (Philippine Mahogany) in Fig. 2 and Annex A. The flashboard will be 12 inches wide.

Consider the bottom board:

Total hydrostatic load, F = 1 920 lbs. (Annex A)
Length (span) L of board, as a beam = 4 feet
Bending moment, as simple beam subjected to a uniformly distributed load, =

Allowable unit bending stress, wood = f_b 6M/bt² = 1 800 psi Solving for thickness, t, inches,

A thickness of 1.78 inches is required of the board against bending. A nominal thickness of 1¾ in. is adequate.

First group lumber gives higher allowable unit bending stresses. When used, this results to a thinner section for a better grade of wood.

Note: Because most lumber distributions still use English measures, the units used are that of the English system, but the figures can easily be convertible to their metric equivalents.