Engineering skills in ancient and medieval Sri Lanka
Modern engineers wondered how the enormous upright monoliths or the slabs of granite used in the anicuts, were quarried, dressed and transported. Bridges, for example, were constructed using stone slabs and columns which were too large for manual handling. Iron was used in the Anuradhapura period. There was plenty of high quality iron. There were 10,000 tons from the Sigiriya site alone. Iron has been used as reinforcement for the Mahathupa, in the time of Dutugemunu.
by Kamalika Pieris
@ The Island - Nov 2007
The ancient Sinhalese were skilled engineers. This can be seen in the works that have come down to us, such as the stupa and the irrigation system. Contemporary engineers have saluted this achievement. The logo of the Institution of Engineers, Sri Lanka (IESL) has as its most prominent feature, a dagoba. Other items in the logo include a sluice gate and a reservoir. This logo was reproduced on a stamp in commemoration of the 100th anniversary of the IESL. The IESL has also included studies on ancient engineering skills in its multi-volume publication ‘History of Engineering in Sri Lanka.’ (2006)
C.R. Panabokke stated at a recent talk before the Royal Asiatic Society, that irrigation in ancient Sri Lanka was an indigenous development. The techniques were not introduced from India. The dry zone was not flat, though it appears so. It is undulating, which meant that it contained valleys and had a lot of drainage. The irrigation schemes of the north central province were set in these valleys. Ancient irrigation started first with channels, then ponds using first and second order streams then came the creation of small tanks and thereafter the large irrigation schemes. .
The techniques used in the ancient irrigation projects have evoked admiration. They were considered to be ‘without parallel elsewhere in the world, when it came to sophistication’. Parker said of Pavatkulam, whose original name is not known, that ‘it is astonishing to find this early work adhering to the best type of later designs. There was a single inlet culvert and two outlets. The designer had enlarged the sectional areas of the inlet and outlet culvert from their entrance to their outlets. Without such enlargement the resulting increased pressure would tend to force the water through the joints of the masonry to the back. This would remove the soil in suspension and eventually lead to failure of the reservoir. The embankment had a total height of 24 ft in the deeper part of the reservoir. The four sluices and the high level culvert under the floor of the southern weir indicated the extent to which the reservoir was utilised."
The calculations made by the ancient irrigation engineers were found to match modern calculations. The Kalawewa spillway meets the modern criteria for a spillway in all respects. Modern sluices and anicuts ended up exactly where the ancient sluices and anicuts were. This happened at Urusita wewa sluice, near Ridiyagama, in the 1960s and Maduru oya in the 1970s. Urusita and Maduru sluices were dated to 5th and 6th centuries, respectively. The sluice barrel at Gantale increased from the up stream end to the down stream end by a factor of seven. This agreed exactly with modern criteria. For earthen embankments, impervious clay was used within the core with semi-pervious material providing the necessary bulk, as in modern designs..
Modern engineers have agreed with ancient solutions. Modern anicuts and modern headworks still use the ancient canals.. Kirindi oya dam was built at an oblique angle of nearly 45 degrees, instead of going across by the shortest possible route. It has now been found that this greatly increased discharging power and the ability to take shocks. At Maduru oya, the brickwork above each conduit is in the form of a corbelled arch. This was done to distribute the load to the walls on either side of the conduit, to reduce the stress. The thick brickwork in the arch acted as a second line of defence against seepage. This was the best possible solution to the materials they had to work with. The water supply of Kalawewa had been insufficient and Mahasena wanted to bring the water of Amban gaga into the valley. This was realised when the Bowatenna diversion of 1976 brought the Amban Ganga water to Kala oya through a tunnel.
P.G.Cooray (1967) pointed out that the ideal site for building a dam is along the geological strike. Only then will the same rock type be continuous along the length of the proposed dam. The beds should also dip upstream of the dam so that the force resulting form the pressure of water and the weight of the dam will be perpendicular to the bedding or foliation planes of the rock. That these ideal conditions were known by the ancient engineers of Ceylon is seen in the countless instances where the dam or bunds of irrigation tanks have been sited along the strike of rock outcrops. "The bunds of Galgamuwa, Sorabora, Minneriya and Kala wewa are only a few of the very many that come to mind."
The ancient engineers were clever. A feature of the Yoda ela is that the canal had only one bund. By doing this, they reduced the cost by half. Another advantage was that with the influx of water, two bunds would have increased the pressure causing damage. With one bund the water spreads on the upper side and releases the pressure so that there is no danger to the bund. Parker admired the design of Panda wewa (near Hettipola.), which he considered to be the oldest of the ancient irrigation reservoirs.
He said ‘owing to the heavy rainfall, in the catchments, the maximum flood could amount to 12,000 to 14,000 cubic feet per second (350 to 400 cubic meters per second). Every engineer will recognise that to get rid of this volume of water in safety would be a problem .The designer of the works must have been a highly intelligent man to overcome it so successfully. The valley had been carefully examined and the embankment laid out making use of an existing rock as a sill for the escape of floods. Beside this, every effort was made to reduce the quantity of earth work to a minimum. This was achieved by aligning the bank to avoid low ground. The fact that the reservoir remained in working order until perhaps the 12th century is proof that the height to which flood would rise over the rock had been correctly estimated.
Present day engineers have noted the difficulties the ancient engineers would have had to deal with. They commented that regulating and discharging at a depth of ten to thirteen metres of water would have been an arduous task in the absence of machinery. They noted the manner in which problems were solved. In some large tanks, wave formations occurred and waves beat the inside of the tank bund. To prevent wave erosion, bunds were strengthened with stone masonry (ralapanawa). The bund could also be destroyed by seepage of water. The ancient engineers had prevented this by digging a trench at suitable places horizontal to the bund and filling it with clay, mostly from ant hills, or kirimati (puddle clay). They got elephants to trample the clay into a thick mass effectively preventing any seepage. The outside area of the tank bund and some other areas were covered with grass to prevent damage by heavy rain. Since the top of the bund was flat the sides had a gradient of less than 45 degrees as a preventive measure.
Since the volume of water in the large tanks exerted considerable pressure, the Sinhala engineers designed the bisokotuwa or cistern sluice. The bisokotuwa is original to Sri Lanka .Brohier described this sluice. "The water from the reservoir flows into the bisokotuwa with very high pressure through strong tubes made of stone. These are generally 2-4 feet wide. The bottom of the bisokotuwa into which this high pressure water flows is generally made of stone. Water loses its pressure as it is made to move upwards and make its exit into a second bisokotuwa where the same process is repeated, reducing the water pressure further. Finally the water, unbelievably disciplined and calm flows out through a larger tube to the outside of the dam.
This tube is normally six feet wide at the beginning and then widens to about 14 feet at the exit point of the dam. That reduced the pressure still further. This was an outstanding idea based on precise mathematics. "
The ancient engineers had superb surveying techniques. Brohier, himself a surveyor, stated that the ancient engineers were highly skilled at surveying. They were able to accurately work out heights and distances. They also knew levelling. The ancient irrigation schemes were on undulating land, though the land looks flat to the naked eye. Brohier noted that irrigation channels were traced mile upon mile, on gradients that would call into use the most precise instruments of the modern age.
"Hence, we must conclude that a system of measuring heights and distances must have attained a very high level of efficiency. Their knowledge included some means of ranging out lines and appraising even small differences in elevation. Such gradients would need very precise instruments even today. But even with modern instruments, it would not be possible to improve on the work of the ancient engineers. They had been so exact" said Brohier. S. de S. Wijesundera said that mapping the gradient for Jaya Ganga or Yoda ela would have been a simple matter for the ancients, judging by the other work executed by them. He pointed out that this project would today be done only after a detailed survey with the aid of measuring instruments. They appear to have possessed appliances almost on par with those in use in modern days. Inscriptions indicate that there was some sort of surveying organisation. .
The engineers had engaged in careful observation. They had studied the topography, rainfall patterns, and soil properties of the island. Site selection was excellent. Parker said of Pavatkulam, ‘the valley had been well explored before the position of the embankment was decided upon. The site was undoubtedly the best one in the whole valley for the formation of a storage reservoir. The engineers would have obtained rainfall figures, to ensure that the rainfall in the catchments area was sufficient to fill the tank. They would also have obtained data on run off and storage. The Jaya ganga was planned because the ancient engineers knew what the levels at Kalawewa and Anuradhapura were. But their approach not confined to practical knowledge. They also had an academic approach to the subject. They had "an amazing grasp of hydrodynamics." They knew the relationship between head and pressure of water and the behaviour of liquids in closed conduits.
Brohier is definite that the ancient irrigation engineers had worked according to design. He thought that plan drawing was practised by second century BC. He said that the engineers had planned beforehand. "It is unlikely that work was started before determining heights and distances. There was also the philosophy which governed the design. the ancient engineers tried to use nature not dominate it. Water was used and reused many times. In the modern system, water is used just once and any excess water is discharged into drainage. The ancient schemes were designed to function as water and soil conservation systems as well as for irrigation.
The engineers experimented and innovated. According to R.A.L.H. Gunawardana there were three types of tank sluices. The smaller tanks had the Keta sorowwa or pipe sluice. Though this is considered a British device, it is actually a copy of the ancient terra cotta keta sorrowwa. The larger tanks had the Rajmohol sorrowwa or piston sluice. This is known from its Indian evidence, but the local evidence is lacking. The largest tanks had the bisokotuwa or cistern sluices. This can still to be seen in various states of preservation in some of the larger unrenovated tanks and even in the occasional small tanks, giving an indication of its versatility. The engineers had also experimented with the use of obtuse angled elbows when releasing water. The Pavatkulam, Periyankulam and Maduru oya reservoirs illustrate this process.
The British during their period of rule over Sri Lanka took the position that the ancient irrigation systems were based on intuitive or practical thinking. ‘The Sinhalese possessed profound practical knowledge of the best methods of dealing with water’ said Parker in 1909. This cannot be accepted today. But it is worth recording that S. de S. Wijesundera remembered an engineer (M.M. Ismail) from Sammanturai, who was involved in the Gal Oya irrigation scheme of the 1950s. He was at ease in the jungle, would never get lost, and had a fine sense of direction, as well a familiarity with even the animal tracks in the jungle. He could walk along a contour keeping to the same level without any assistance or instruments. A field officer would come behind marking the initial canal trace Because of this he could construct canals at great speed.
Because there was little appreciation of their historical importance, the remains of ancient works were often erased when modern schemes were implemented. However some ancient irrigation works can be found undisturbed even today in the forested areas of the dry zone. C.R. Panabokke says that there are excellent remains of ancient tanks in Vilachchiya. No one has looked at them. He also said that the greatest concentration of ancient anicuts was in Matara, where there were about 13,000 anicuts and about 13,000 small tanks.
The ancient engineers knew to design and build large structures, such as stupas and bridges. In doing buildings they first perfected the foundation. they constructed framed structures as well as load bearing and earth compacted structures. They also constructed arches, vaults, and domes. Architect Nimal de Silva said that in most of the major works, the systems used were unique and indigenous.
Structural engineers have looked at the structure of the stupa. The stupa had a solid foundation, usually on rock. The dome was constructed either of brick, or was earth filled, with outer brick wall. Jetavana had full bricks, half bricks and an earth fill. All ancient stupas had been initially in the shape of a paddy heap. This shape was good for stress. Abhayagiri was a perfect parabaloid and Jetavana was a perfect ellipsoid. Demala maha saya was a perfect circle.. There were perfect verticals. Even after the dome had collapsed, the spires of Abhayagiri and Jetavana stood perpendicular to earth. Loha maha pasadaya would have been triangular, which means that calculations regarding the distribution of weight in each set of pillars would have been necessary. Nimal de Silva observed that the perfect linear arrangements seen in the ancient cities showed that engineering plans had been prepared. construction work on the Mahathupa had been supervised.. .Labour was paid for and. various necessities, such as food and clothing were provided at the workplace.
The builders used the standard materials of the time, stone, timber and clay. P.G.Cooray pointed out that the clay deposits in Sri Lanka are extensive and they occur in the flood plains of the major rivers in old river channels, owitas, in tank beds, and in the deep weathered zones of the crystalline rocks. Some of these deposits had been utilised for centuries. the clay and sand for Ruvanvelisaya were found in vicinity In Malvatu Oya basin. Bricks for Mahathupa (Ruvanvelisaya) were made of selected clay and the work was closely supervised. Two monks had on separate occasions offered bricks which did not conform to the specifications and the bricks were rejected. This means they exercised quality control.
Iron was used in the Anuradhapura period. There was plenty of high quality iron. There were 10,000 tons from the Sigiriya site alone. Iron has been used as reinforcement for the Mahathupa, in the time of Dutugemunu. Wijesundera commented ‘the network of iron on Mahathupa may not be the type of mesh that we now have or even round bars. in all probability, they would have been rectangular bars in short lengths interlaced in either direction." During the time of Parakrama bahu II (1153-86) there were carriages drawn by horses, where iron bands were used. The first use of steel as reinforcement in concrete, in the west had been in France in 1968."
There were stone bridges over Kala oya, Malvatu and Yoda ela. At Kala oya, Forbes, had found the remains of a bridge, over a stream which was broad and rapid. Stones used were from 8 to 14 feet in length. They were laid in regular lines, some jointed into one another. each course receded a few inches from the edge of the one underneath. this while it offered less direct resistance to the current, give additional strength. The rock contained square holes, in which stone pillars had been placed. the bridge had been completed by laying long stones or beams of wood on these stones to connect the different parts of the structure. On most bridges the pillars were set above one another using a stone collar. The posts were in two sections totalling 18 feet in height, the collar stones were 3 feet, roughly squared or rounded. The Bogoda bridge (14th century) did not use nails.
Modern engineers wondered how the enormous upright monoliths or the slabs of granite used in the anicuts, were quarried, dressed and transported. Bridges, for example, were constructed using stone slabs and columns which were too large for manual handling. Brohier said that the stones had been extracted from the adjacent rocks by means of wedges and then shaped by chisels. Forbes writing earlier, noted that that the method of working quarries and splitting stones was "every where observable." He said that these techniques were introduced into Britain only in the 19 century.
The ancient Sinhalese used machinery. Mahavamsa says Bhatika Abaya raised water by means of machines from the Abhaya tank. Siriweera says that sugarcane juice was extracted by a machine, the 'uk yantra'. Modern engineers also think that the ancient engineers had machinery. They say pulleys and levers would have been used for erecting columns. Holes had been made in the base rock to receive the columns into position. Scaffolding or platforms may have been used. They had very good tools. The brahmi inscriptions were cut on slabs of hard granite. These inscriptions had regular lines and smooth surfaces showing that chisels of superior quality steel had been used. Europe had no steel that could cope with the local gneiss. They also had very good survey instruments. The stupas were precisely oriented along the north–south, east-west axes varying within one degree. Forbes noted that the ancient engineers have had "appliances almost on par with those in use in modern days."
The writings of R.L.Brohier, P.G. Cooray, Nimal de Silva, P.L.N. de Silva, S Deraniyagala, W. Geiger, A. Guruge, D.L.O. Mendis, C.R.Panabokke, S. M.P. Ranaweera, M.P. Ranaweera, W.I..Siriweera, P. Sivasegaram and S de S Wijesundera were used for this essay.