York Sensors uses optical fibres to measure localised temperature changes. The system can be used to measure leaks in water-retaining structures and heat of hydration, and was recently used in its first applications in dams and hydro channels

Fibre optic temperature sensing systems have been under development for most of the last two decades, and their application to water-retaining structures has been considered for most of that time. The technology was not employed on dams and embankments, however, until the 1990s. One system, developed by UK-based York Sensors, is now being used by Germany-based GTC Kappelmeyer at four sites: detecting leakages in water and navigation channels and at the Ohra dam; and to monitor concrete curing temperatures at the Birecik dam in

Turkey (above).

The principle of the temperature sensing system is well known. An optical pulse is injected by laser at one end of an optical fibre. The light is back-scattered, and in the process its frequency is shifted by an amount which depends on the temperature at the point of scattering (this is known as Raman shifting). The position of the light scattering point is determined by measuring the delay between the time the light pulse is introduced and the time the back scattered light returns, while the temperature of the point is determined by calculating the amount of frequency shift.

According to York Sensors, manufacturer of the distributed temperature sensor (DTS), its current system, known as DTS800, allows temperature to be measured continuously over fibre lengths of up to 30km, and offers a temperature resolution of ±0.2°C and a sampling resolution of ±0.25m. The system can be prepared as a single end or double end installation — the latter allows for a system integrity check and therefore greater accuracy. Experience of the system has previously included power cables, chemical process industries, and fire detection systems in tunnels and mines (more details are given on York’s website at www.york-sensors.co.uk). In the hydro industry the system has caused most interest in the area of leak detection, as leaks are generally characterised by a local change in temperature.

Leak testing

York’s temperature measurement system was first applied to dam construction in December 1996, in a research programme carried out for Hitachi at the Munich University of Technology. A system similar to a 2m high zoned embankment dam with core sealing was built and more than 300m of fibre was laid in loops at different depths, along with 50 conventional temperature sensors. Within the clay core a leak comprising a sand and gravel window was introduced.

  When the system was impounded, by filling with water at 0.2°C in the upstream shoulder, the leak was located inside the warmer fill of the downstream shoulder. Following the success of the first test the system was installed during 1997 at two facilities.

Application: Mittlerer Isarkanal

The DTS system was first used for surveillance of a hydro supply channel at the Mittlerer Isarkanal near Munich, Germany. An existing channel at this site was being refurbished, and during the rehabilitation of the sealing system two fibre optic cables were deployed in a ditch filled with drainage gravel along the length of the channel. The cable was embedded in fine gravel about 1.2m underneath the facing in two sections. One was underneath a 200m section with concrete facing and the other was underneath a 1100m section with asphalt facing. The cable was deployed in August 1997; measurements were made daily during the flooding of the channel and weekly for the following two months. Measurements are planned twice-annually in future.

Results show that both sealing systems are functioning well. Some effects were observed:

•In the first two weeks high permeability of the concrete sealing was seen, probably from open construction joints.

•A temperature drop, caused by a small leak, was detected in the asphalt section. The leak was sealed after a few weeks by fine sediments in the channel water and the temperature returned to normal.

•In the concrete section, at the upstream toe of one of the embankment dams, local seepage zones were observed some three months after the flooding. Using the temperature values obtained through the optical fibre it was possible to clarify that the seepage is due not to a defect in the sealing system but to groundwater flow reinforced by ground works during the channel rehabilitation. Thanks to this information the cost of improving the new sealing system was not incurred.

Application: Birecik dam

The DTS800 system is now being used at the Birecik gravity dam, currently under construction at the Euphrates river in Turkey. The system was deployed within the dam structure during concreting and will provide near-continuous temperature measurement during the period of concrete curing. The system was incorporated into the dam as a 155m long cable inside the construction block, in the summer of 1997. The cable has been placed in three zones of block 4A, near the main bottom gallery of the dam.

The installation was carried out by fixing the cable on a rack formed by reinforcing bars, which cross the whole block from the upstream side to block 4B on the downstream side. Block B had already been built. The optoelectronics measure-ment unit was installed in the bottom gallery. Measurements were carried out during the first two weeks after concreting, and additionally five months later.

The heat of hydration was measured continuously, and across the whole block during the first two-week long, measurement period. A maximum concrete temperature of 38°C was reached after around one week, by which time the temperature distribution across the block was roughly constant and depended on the local pouring temperature only (see diagram right). When measured five months later the concrete temperatures were defined by their distance from the upstream side of the block.

The temperature distribution obtained from the optical fibre showed that the cooling front is advancing within the block, thanks to the combined effect of the seasonal decline in air temperature and the heat of hydration. The downstream part of the block — at the greatest distance from the free surface — had remained at the maximum concrete temperature of 38°C for five months, except for regions cooled by their proximity to the gallery.

The more-or-less continuous data obtained across the block from the fibre optics were compared with results from conventional thermometers, and showed good agreement.