The Solis dam was built in 1986 as part of a cascade scheme in the Swiss Alps. By 2012, its total storage volume had been reduced by about 50 per cent due to sedimentation. A sediment bypass tunnel was built to reduce the sedimentation rate and maintain the storage volume.
The Electric Power Company of Zurich (ewz) built the Solis arch dam in 1986 to impound a 3 km-long reservoir with a storage volume of 4.1 Mm3. The Solis dam is the third dam in a cascade scheme in the canton of Grisons in the Swiss Alps, as shown in figure 1. The Solis reservoir receives water from the Julia and the Albula Rivers, plus turbined water from the 52 MW Tiefencastel East and the 24 MW Tiefencastel West power plants immediately upstream.
While the upstream Albula River basin remains natural, hydropower potential in the Julia River basin has been harnessed with two dams, the Marmorera and the Burvagn reservoirs, and four power plants. The Solis reservoir supplies water to the 26 MW Sils and the 38 MW Rothenbrunnen power plants further downstream. Water storage capacity loss at Solis would decrease the reliability of energy production in the power plants downstream.
The reservoir created by the arch dam, which is 61 m high with a crest length of 75 m, originally had a total volume of 4.1 Mm3, of which 1.5 Mm3 was live storage. The maximum and minimum operating levels are 823.75 and 816 masl respectively. The catchment tributary to the Solis dam comprises 900 km2 and contributes about 110,000 m3 of sediment each year. Due to the Alpine catchment characteristics, the Julia and Albula Rivers convey the majority of sediment during the snowmelt period between May and July, and during the heavy rain periods between June and September.
Sediment is trapped in the Solis reservoir and deposited in the topset reach, forming a long delta with a characteristic pivot point.
The successive longitudinal profiles obtained from echo-sounding surveys in figure 2 show the delta moving steadily forward towards the dam. Because the range of operating levels has been stable, the delta exhibited little vertical growth; most incoming sediment was being deposited at the delta face, moving it downstream toward the dam. Finer sediment was being deposited downstream of the delta. Bottom outlet flushing in 2006 and 2008 was carried out exceptionally to lower the sediment body. From 2009, flushing was no longer permitted.
With a long-term annual average rate of storage loss of around 80,000 m3, sedimentation had caused a 1 million m3 decrease in the reservoir's live storage by 2009. By 2012, following 25 years of sedimentation, the reservoir had lost half of its original total storage capacity.
Without sediment management measures, the advancing coarse delta sediments could have potentially resulted in the clogging of the bottom outlets. This would have led to a plant shutdown in the long term.
Sediment management strategies
Before taking the decision to build a sediment bypass tunnel, ewz considered other sediment management strategies. The company considered feasibility studies of strategies including dredging, flushing and sluicing. Dredging was excluded as an option for ecological and economic reasons.
Empty flushing was discarded for financial and technical reasons. These included the income loss associated with plant shutdown during the flushing activity, the low capacity of the bottom outlets, and the technical challenge of completely emptying the reservoir through the existing outlets before the arrival of the flood, given that this approach would not allow for the drawing down of the reservoir completely before a flood event in the small steep watershed.
Two options based on partial reservoir drawdown were tested through physical modelling at ETH Zurich, Switzerland: partial drawdown sluicing through the existing low-level outlets, and partial drawdown flushing through a sediment bypass tunnel (SBT). Both were found to be technically effective in stabilising the delta by exposing it to scouring action during floods and retaining the live storage volume. However, the bottom outlets that would be used for sluicing have very low capacities to convey the high suspended sediment load and, if blocked by driftwood, it could threaten the dam’s operational safety, requiring prolonged shutdown to remove the blockage.
The Solis reservoir acts as an important power hub in the region, and if taken out of service it would affect power production at a number of other plants. For this reason, it was particularly important to minimise the risk of plant shutdown.
The cost of sluicing was estimated at USD 2 m for initial investment, plus USD 0.35 m in annual maintenance, while the cost of the SBT was USD 38 m plus USD 0.90 m for annual maintenance. The main drawback to the SBT was the high financial cost. A risk analysis was carried out to compare the two options. The risk of plant shutdown due to outlet clogging under the sluicing scenario was converted into an equivalent financial cost, and the two projects were compared on the basis of actual cost plus an imputed cost of risk. Despite the higher initial cost of the SBT, when the cost for risk was assessed in the present value analysis, the performance of the SBT was comparable to sluicing.
The SBT was ultimately chosen as the most effective strategy to maintain the active storage capacity of Solis reservoir. In addition, the SBT was considered to have downstream environmental benefits because it would more effectively pass coarse sediment bed material below the dam, offsetting downstream streambed degradation due to coarse sediment trapping in the reservoir.
The Solis bypass tunnel was designed with an archway cross-section of 4.40 m by 4.68 m and the tunnel invert is lined with high-performance concrete. The grain size of the bypassed sediment is 6 cm for D50 and 15 cm for D90 and the sediment material has a low quartz content. The tunnel is 968 m long with a steady slope of 1.9 per cent. The tunnel flow is supercritical and only uniform at the end. The maximum bypass flow velocity is 13 m/s, the design discharge is 170 m3/s and the minimum discharge is 90 m3/s.
Without sediment management measures, the advancing coarse delta sediments could have potentially resulted in the clogging of the bottom outlets."
The general layout of the SBT is shown in figure 3. When the reservoir is partially drawn down, water is diverted into the tunnel entrance by a 140 m long guiding structure that crosses the river section, which is submerged during normal reservoir operations. Physical modelling of the SBT showed that the tunnel would capture and bypass all the sediment, while only large floods that exceed the tunnel capacity and overtop the guide wall would carry suspended material further downstream. Bed load is entirely diverted into the tunnel even if the design capacity is exceeded. A submerged skimming wall was also provided to prevent driftwood from entering the tunnel. The skimming wall diverts floating debris towards the dam where is mechanically removed or spilled. The tunnel was lined with high-strength concrete and the SBT outlet is a cantilever structure that discharges about 12 m above the Albula river.
The Solis bypass tunnel differs from the usual bypass designs in the intake location. Rather than being located at the reservoir head, the intake is 450 m upstream of the dam. This design has the advantage of reducing the tunnel length and avoiding the need to create an acceleration section because the tunnel entrance will be under pressurised conditions. This location also allows the face of the delta to be intercepted. The upstream sedimentation body is lowered down to the desired level according to the partial draw down, and the downstream part is unaffected. However, two things are complex in this design: (1) the intake level has to be specified according to the desired sediment 'free' level, which would preferably be the minimum operating level; and (2) the operation becomes more complex because if the partial drawdown is not enough, the sediments are not transported, but if the partial drawdown is too low, sediment deposits are mobilised. These differences are highlighted in figure 4.
Sediment bypass tunnel operations
The general operational elements and corresponding hydraulic conditions in the Solis reservoir are summarised in the following points and figure 5.
In anticipation of high inflow events, the reservoir is drawn down to the minimum operating level of 816 m. This creates scouring flow across the top of the delta and also ensures that the sediment-laden flow is directed into the tunnel intake by the guide wall.
Drawdown typically requires 8-10 hours. Refill time is shorter, typically about five hours.
Power is normally shut off during a large flood to protect the machinery from sediment abrasion, and by timing the drawdown and bypass event to coincide with floods, the impact on power production is minimised.
The bottom outlet is normally opened to control water levels during periods of high inflow. The bottom outlet at the dam can be opened for pressure flushing and venting of density currents at any time with the reservoir at a high level, since this procedure only releases small volumes of sediment. Under the ideal operation, the bottom outlet starts closing after starting the SBT operation.
Authorities require >70 m3/s for flushing. This requirement minimises downstream environmental impacts as the large flow will help transport the sediment downstream.
Clear water is released at the end of the flushing event to clean the riverbed below the dam.
The sediment released through the bypass tunnel subsequently passes along the downstream channel as a sediment wave.
The challenge for the operators is the need to perform the operation with limited data. The operation is complex and requires a lot of coordination within the company and with downstream authorities. The energy trader, in particular, will want to know exactly when power will be taken offline.
Operational personnel were involved in the project design to ensure that the final design met the operational needs of the plant.
The first operations of the bypass tunnel took place in May 2013. The maximum discharge flow was 110 m3/s and the bypass lasted 12 hours. All inflowing sediment was diverted through the tunnel and the top of the delta was scoured. In August 2014, a major flood with a 100-year return period and a peak discharge of 288 m3/s occurred. Over 14 hours, the flood was routed through the tunnel achieving a maximum discharge of 179 m3/s. The excess flow overpassed the guiding structure and as a result, the aggradation in the reservoir increased by 102,000 m3. However, this volume represents less than half the deposit volume that resulted from a comparable event in 1987. The sediment bypass tunnel, therefore, proved to be effective for both moderate and large floods.
The bypass tunnel has worked as planned. The positive effect in controlling delta growth and sustaining live storage volume can be deduced from the sediment profile given by the echo-sound survey in 2013. About two thirds of the annual sediment load is now bypassed through the tunnel.
In addition to real-time hydrologic monitoring in the watershed as required to predict flood inflow, sediment concentrations are measured for ecological reasons both upstream and downstream of the reservoir. Both suspended and bed load are monitored. In 2015, new monitoring systems were installed to measure bedload sediment transport. The monitoring system was located 1.5 km upstream of the reservoir in the Albula River and it consists of 15 accelerometer sensors together with 15 geophone sensors.
The abrasion rate on the tunnel floor is monitored by repeated laser scans, and was found not to be significant after the first bypass operations. Several test sections of different types of floor materials were also installed and are being monitored for abrasion.
A further scientific study (Auel et al. 2017) suggests the positive effect of long-term SBT operation on the microhabitat of the downstream ecosystem. It revealed that the microhabitat and invertebrate richness increased downstream of the SBT outlet the longer the SBT was in operation.
Graphs and figures
Auel, C. and Boes, R. 2011. Sediment bypass tunnel design - review and outlook. Dams and Reservoirs under Changing Challenges. Editors: Schleiss and Boes. Taylor & Francis Group, London, ISBN 978-0-415-68267-1
Auel, C. et al 2017. Effects of sediment bypass tunnels on grain size distribution and benthic habitats in regulated rivers. International Journal River Basin Management.