Abstract
This paper presents the emergency response to the barrier lake created by the Mw 6.5 earthquake on August 3, 2014, in Niulan River which is situated in Yunnan Province, southwest China. The lake, behind 83- to 103-m-high landslide dam, has a storage capacity of 260 million m3. The Hongshiyan hydropower station, situated just upstream of the landslide dam, was submerged by the lake water. The rescue actions include building an 8-m-deep drainage channel in the flood season, draining the lake water via a water release tunnel and rebuilding the landslide dam as a permanent structure. This paper presents the considerations and processes in deciding on the emergency actions. One major consideration is to minimize the hazards caused by a possible dam break flood induced by Hongshiyan barrier lake in the flood season. It was found that the barrier lake is classified as “grade I: extremely high risk” in which the 5- and 20-year floods should be used for design and check purposes of emergency works, respectively. The dam break flood evaluations indicated that an 8-m-deep diversion channel should be constructed in order to reduce the total amount of the released reservoir storage and to prevent possible overtopping of the downstream Huangjiaoshu rockfill dam. This study was based on an improved analytical approach to dam break flood modeling and a handy spread sheet developed by the authors. The considerations on various risks and the key analytical technology for dam break flood predictions described in this paper may be useful for other similar cases.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
1 Introduction
On August 3, 2014, an earthquake of magnitude 6.5 occurred at Ludian, Yunnan Province, in southwest China. As a consequence of the earthquake, a barrier lake was formed in the Niulan River. This river is a tributary on the right bank of the Jinsha River. The landslide debris formed an 83- to 103-m-high barrier dam, which can hold a maximum water storage of 260 million m3. The landslide is located 600 m downstream of Hongshiyan hydropower dam that was completely submerged by the rising water of the barrier lake. Downstream of the landslide dam, there exist a 90-m-high concrete arch dam and a 65-m-high concrete face rockfill dam, which are located at 18.8 and 57.6 km from the landslide dam, respectively. Break of the landslide dam not only poses a threat to the downstream residents but also has a risk of overtopping the two downstream existing dams. This barrier lake also has a special feature in that once the impounding pool water reaches the elevation of 1174 m (which is approximately 48 m below the dam crest), the water stops rising under normal conditions because the water level reaches the elevation of the surge tank crest of the submerged Hongshiyan hydropower station. Thus, the water in the river flows through the water supply tunnel, overflows the surge tank and then flows downstream. This means that the dam will stay unless it is destabilized by piping or overtopping with a huge flood whose flowrate is much greater than the discharge capacity of the water supply tunnel. In view of the situation, it is necessary to ask the following questions:
-
1.
What is the dam break flood if ever the dam is overtopped due to an extraordinary huge flood in the Niulan River?
-
2.
What is the flood standard to be adopted for the dam break flood analysis?
-
3.
If the predicted dam break flood exceeds the mitigation standards, what are the possible engineering measures that could lower the flood peak to a level that is safe for the downstream residents and structures?
Occurrences of landslide dams are common. Strong earthquakes are a main triggering factor (Keefer 1984, 2002; Evans et al. 2011). Typical examples of earthquake-induced barrier lakes can be referred to the Tsao-Ling landslide dam formed by Chi-Chi earthquake on September 21, 1999, in Taiwan (M w 7.6; Li et al. 2002; Khazai and Sitar 2003), the EI Desagüe and Jiboa landslide dams in El Salvador on February 13, 2001 (M w 6.6; Baum et al. 2001), the Higashitakezawa barrier lake formed by Niigata-ken Chuetsu earthquake on October 23, 2004, in Japan (M w 6.6; Furumura and Hayakawa 2007), HattianBala landslide dam formed by Kashmir earthquake on October 8, 2008, in Pakistan (M w 7.6; Sato et al. 2007; Schneider et al. 2013). In China, the M w 8.0 Wenchuan earthquake triggered approximately 60,000 landslides (Dai et al. 2011; Gorum et al. 2011) and formed more than 800 landslide dams (Fan et al. 2012). The Tangjiashan barrier lake had the highest risk among them, and fortunately, it was successfully drained in time (Liu et al. 2010).
The experts involved in this emergency response work were mainly the key members who successfully carried out the rescue work of Tangjiashan barrier lake. They carefully reviewed all the related concerns and came to the following conclusions that were put into action:
-
1.
According to the code “Standard for Classification of Risk Grade of Landslide Lake (SL 450-2009)” published by the Ministry of Water Resources of China, the Hongshiyan barrier lake belongs to risk grade I and the 5- and 20-year floods should be used for design and check purposes of emergency works, respectively, because the rescue work was carried out during the flood season of this river.
-
2.
The dam break analysis indicates a flood with a peak discharge of 8368 m3/s associated with the 20-year flood. This would cause overtopping of both Tianhuaban and Huangjiaoshu dams.
-
3.
A decision was made that an 8-m-deep diversion channel be constructed in the right embankment in order to reduce the total amount of the released reservoir storage by 1.2 × 107 m3 and the peak dam break discharge to 8163 m3/s. This measure would ensure that the possible dam break flood would not overtop Huangjiaoshu rockfill dam.
-
4.
As a matter of fact, the 20-year flood did not occur in the 2014 flood season and the diversion channel was not used. The reservoir was subsequently emptied by a new diversion tunnel. This landslide dam is being upgraded to a permanent dam replacing the submerged Hongshiyan concrete dam.
These emergency actions are considered appropriate in view of the huge disaster that could happen if the dam were to break. As members of the working team on the dam flood evaluation for the Headquarters of Hongshiyan Emergency Action, authors of this paper present the various aspects that were studied during the emergency response period which led to the final hazard reduction decisions.
2 Characteristics of river basin and Hongshiyan barrier lake
2.1 Hydrology of the Niulan River
The Niulan River, where the landslide dam happened, is a main tributary of the Jinsha River. It is located in Yunnan Province, China, as shown in Fig. 1. Its total length is 440 km with a catchment area of 13,672 km2. The average bed slope is approximately 4.4 ‰. According to the historical data from 1953 to 2012, the average annual runoff was approximately 4.9 × 109 m3, and the average flow at Hongshiyan hydropower dam was approximately 128 m3/s. In the typical year of 1968, the peak discharge at Hongshiyan hydropower station was measured to be 2740 m3/s. Figure 2 shows the design flood hydrographs of different return periods (solid lines) and the 1968 flood hydrograph (dashed line).
The Niulan River has rich water resources with four hydropower stations that are located in the vicinity of the landslide. Figure 3 shows the locations of the four cascade hydropower stations and the landslide dam. The Hongshiyan hydropower station is submerged completely, and the downstream hydropower stations (Tianhuaban and Huangjiaoshu) face the risk of overtopping if the landslide dam breaks. The main technical parameters of the downstream hydropower stations are shown in Table 1.
2.2 Hydropower projects
2.2.1 Hongshiyan hydropower station
The Hongshiyan hydropower station has its dam located just 600 m upstream of the landslide and was completely submerged by the barrier lake. The Xiaoyantou project has a concrete arch dam located 29.6 km upstream of the barrier lake, which was not damaged during the earthquake. Hence, it was not part of the rescue work for Hongshiyan barrier lake. The project has a water supply tunnel in the right abutment that leads to the power station. Figure 4 shows the layout along the diversion channel of Hongshiyan station. After the earthquake, in the afternoon of August 4, 2014, the water in the surge tank of the Hongshiyan station started to overflow (refer to Fig. 5). This happened when the water level rose to 1171.80 m, which is the elevation of the surge tank crest. The relationship between the overflow discharge of the surge tank and the bulkhead gate and the water level of the lake is presented in Fig. 6. However, there is still a risk of the landslide dam to be overtopped if the river encounters a huge flood in the flood season. Figure 7 shows the relationship between the water storage and the elevation of the water level in the landslide lake.
2.2.2 Tianhuaban hydropower station
The water-retaining structure of Tianhuaban hydropower station is a concrete arch dam. The dam is 107 m high. The elevation and the length of the dam crest are 1076.8 and 159.87 m, respectively. The discharge structures include three surface spillways, two middle orifices and a sediment bottom sluice. The inlet elevations of surface spillway and middle orifice are 1260.5 and 1020 m, respectively. The elevations of the check and normal flood levels are 1076.61 and 1070.91 m, corresponding to the reservoir capacities of 7.87 × 107 and 6.57 × 107 m3, respectively. The maximum discharge is approximately 5046 m3/s through all discharge structures, and the discharge would be 3152 m3/s if the water level is at the design flood level (1070.91 m). The storage–elevation curve and the discharge curves are shown in Figs. 8 and 9, respectively.
2.2.3 Huangjiaoshu hydropower station
Downstream of Tianhuaban hydropower project, there is a concrete face rockfill dam for the Huangjiaoshu hydropower station. The safety of this dam is of prime concern because the consequence of this dam being overtopped due to the breaking of the barrier lake is unacceptable. The dam is 65 m high. The elevation and the length of the dam crest are 775 and 217.72 m, respectively. The elevations of the check and normal water levels are 774.37 and 770 m, corresponding to the reservoir capacities of 4.05 × 107 and 3.29 × 107 m3, respectively. The discharge structures include a spillway tunnel and a sediment diversion tunnel. The respective inlet elevations of the two tunnels are 750 and 716 m. The maximum discharge is approximately 6691 m3/s through both tunnels, and the discharge would be 5744 m3/s if the water level is at the design flood level (i.e., 770 m). The storage–elevation curve and discharge curves are shown in Figs. 10 and 11, respectively.
3 The Hongshiyan barrier lake
3.1 Hongshiyan landslide dam
Hongshiyan landslide dam is formed by the collapsed strongly weathered dolomite on both sides of the Niulan River. The volume of the landslide deposit is approximately 1.2 × 107 m3 with the deposit on the right of the landslide accounting for 70 % of the total. Some of the deposits in the dam body are larger than 50 cm, and the maximum diameter of rock is between 5 and 10 m. The crest elevation of Hongshiyan landslide dam is 1222 m. Its height is between 83 and 103 m. The storage capacity of barrier lake is 2.6 × 108 m3, and the length of upstream backwater is 25 km when the water level is at the crest elevation of landslide dam (i.e., 1222 m). Table 2 and Fig. 12 show the main characteristic parameters and topography of the landslide dam.
3.2 Assessment of the level for risk control
The risk classification of a landslide dam depends on the hazard level and the severity of the consequential loss. The Ministry of Water Resources of China published a code entitled “Standard for classification of risk grade of landslide lake (SL 450-2009).” According to the size of barrier lake, a dam can be classified into one of the four categories: “large” (V ≥ 1.0 × 108 m3), “medium” (1.0 × 108 m3 > V ≥ 0.1 × 108 m3), “small (1)” (1.0 × 108 m3 > V ≥ 0.1 × 108 m3) and “small (2)” (V < 0.01 × 108 m3), where V is the water storage of the lake. Further, according to the size, material composition and the height (designated H), the risk of landslide dam can be classified into “very high” (H > 70 m), “high” (30 m > H ≥ 70 m), “significant” (15 m > H ≥ 30 m) and “low” (H < 15 m). The severity of consequential loss due to breaking of landslide dam is directly dependent on the number of people at risk (PAR), and the importance and values of the towns and its installations. It can be classified into four levels: “extremely serious” (PAR ≥ 106), “serious” (105 > PAR ≥ 106), “relatively serious” (104 > PAR ≥ 103) and “general” (PAR < 104).
The risk of landslide lake can be classified into four categories based on the hazard level and the severity of consequential loss in the case of a dam break. Using the risk matrix method, Table 3 shows the risk classifications.
3.3 Selection of design flood for emergency mitigation civil work
For any structure that is to be constructed on a river, it is a common practice to design it based on a flood of a certain return period (Fernández and Salas 1999). Similarly, the code specifies flood return periods for landslide lakes associated with their risk classifications, which are shown in Table 4.
Based on Table 3, the Hongshiyan barrier lake belongs to grade I risk classification. Based on Table 4, the corresponding design flood for grade I risk classification is associated with a return period “≥5 year.” However, the code did not clearly specify what a particular year of return period is suitable for a specific grade I barrier lake. After a careful study, it was decided that for Hongshiyan barrier lake, the 5- and 20-year floods should be used as design and check cases, respectively. The main reasons are as follows:
-
1.
From Table 3, one can find that under the “grade I” risk classification, there are two subgrades which are “serious” and “extremely serious.” In view of the possible impact on the two existing dams downstream, Hongshiyan barrier lake is considered to belong to the subgrade “extremely serious” and the upper bound of the specification of “≥5 year” should be adopted.
-
2.
A back analysis for a similar emergency disposal work on the design flood return period for the Tangjiashan Barrier Lake has been recently carried out by the authors’ research group (Zhou et al. 2015). It was concluded that the overtopping of this landslide dam in conjunction with a 20-year flood would create a flood of 17,583 m3/s at Mianyang, a downstream very important city with 1.2 million residents, while the dike of this city can only resist a flood of 15,600 m3/s. A 13-m-deep diversion channel was actually excavated. The back analysis showed that this work would reduce the dam breach flood at Mianyang to a magnitude of 15,464 m3/s for an overtopping combined with a 20-year flood. This case history represents an experience of engineering practice that adopts a flood standard of one in 20-year flood.
After determining the design and check floods, the next question is to select an appropriate emergency mitigation civil work that can meet the requirement of these standards. The feasibility studies include the alternatives of excavating a diversion channel with depth of 8, 14 and 22 m, respectively. It was finally concluded that an 8-m-deep channel would make the downstream Huangjiaoshu rockfill dam survive without being overtopped by the landslide dam breach combined with a 20-year flood. The remaining part of this paper is mainly focused on the alternative with an 8-m-deep diversion channel.
4 Dam break and flood routing analysis
4.1 Method of analysis
Predicting the dam break flood due to overtopping of the Hongshiyan landslide dam, either in its original geometry or in the alternative geometry with an 8-m-deep diversion channel, is an essential part of the emergency mitigation study. There are two parts in this study: (1) estimation of flood hydrographs at the breach and (2) routing of the flood hydrographs along the downstream river until they reach Huangjiaoshu rockfill dam.
There are many publications on the numerical simulation of the dam break process (e.g., Fread 1984, 1988; Singh et al. 1988) and the associated computer programs, such as DAMBRK, BREACH (Fread 1984), BEED (Singh et al. 1988), MIKE11 (Havnø et al. 1995) and HEC-RAS (HEC 2006a, b). However, Wahl (2004) and Zhu et al. (2004) reported that there are considerable uncertainties in the predicted peak discharge. For example, for the possible dam break of the Tangjiashan barrier lake, the predicted peak discharge was 46,000 m3/s, which was eight times larger than the actual measured discharge (Liu et al. 2009). A simplified dam breach model was recently developed by the authors’ research group that is highlighted as follows (Chen et al. 2015).
By equating the discharge through the breach based on the theory of flow over a broad crested weir, and the loss of storage in the reservoir in a unit time, the governing equation for a dam break analytical model is
where C is the discharge coefficient, B is the width of the weir, H is the elevation of water level, z is the elevation of breach bed, q is the inflow of the river, W is reservoir water storage, and t is time.
Both B and z vary throughout the dam break process due to the lateral enlargement and soil erosion of the breach. With respect to the soil erosion rate that relates z to the shear stress τ, Chen et al. (2015) recommended a hyperbolic model:
where k is a unit conversion factor that allows E to approach its asymptote E ult within the working range of \(\tau\). As shown in Fig. 13, the hyperbolic curve has an asymptote represented by E ult = 1/b as \(\tau - \tau_{c}\) approaches infinity. This is the maximum possible soil erosion rate. The parameter 1/a represents the tangent of this curve at \(\tau { = }\tau_{c}\).
Slope stability analysis using a wedge failure has been widely adopted to find the enlarged value of B in Eq. (1) (e.g., Fread 1984). It was found that the breach lateral enlargement process ought to be modeled using more rigorous analytical methods with circular slip surfaces, such as Bishop’s simplified method (1955). The procedure for calculating the factor of safety F is repeated among various possible slip surfaces until a critical one associated with the minimum factor of safety F m is found (Chen et al. 2015). This algorithm enables fast and robust calculations to be carried out using an Excel spreadsheet DB-IWHR 2014, which is available for download at the following website: http://www.geoeng.iwhr.com/geoeng/download.htm.
Since the Hongshiyan dam material is very similar to that of Tangjiashan, all the related parameters for the latter are adopted for the current study, which can be found in Chen et al. (2015).
A dam break flood travels downstream can be mathematically described by the unsteady, open channel flow based on Saint–Venant equations (Fread 1988; HEC 2006b; Moussa and Bocquillon 1996). Based on the analysis of the dam break at Tangjiashan barrier, it was shown that the breaching hydrograph remains practically unchanged in a narrow valley (Liu et al. 2009). Since the Hongshiyan landslide dam and the Niulan River are also situated within a narrow valley, the hydrograph that passes the two dams downstream is therefore expected to be the same as the breaching hydrograph.
4.2 Flood analysis at Hongshiyan barrier lake
It is proposed that flood regulation comes into operation when the water level reaches 1180 m. This is the water level elevation in which the natural flow equals the discharge capacity of the surge tank. The objective of this study is to determine whether the 5- and 20-year floods would cause the lake water level to rise above 1222 m which is the elevation of the dam crest, or above 1214 m which is the new elevation of the dam crest if the 8-m-deep channel is constructed. The process of flood regulation can be calculated by the Eqs. (3) and (4):
in which V is a function of Z
where Δt is time step; Q 1 and Q 2 are the initial and final inflows of calculated period, respectively; q 1 and q 2 are the initial and final outflows of calculated period, respectively; V 1 and V 2 are the initial and final volume of calculated period, respectively; Z is the elevation of barrier lake. Integration is made with the given time step.
The inputs involved in Eq. (3) are as follows:
-
1.
Q 1, Q 2 can be found in Fig. 14a for either the 5- or 20-year flood;
-
2.
q 1, q 2 are the discharge capacities of the surge tank and bulkhead gate, respectively, associated with the lake water level elevation, as shown in Fig. 6, and
-
3.
the relationship between the lake water level Z and the storage in the lake V is shown in Fig. 7.
The flood routing results show that (Fig. 14b):
-
1.
for the 5-year design flood, the water level reaches 1214 m at 32.80 h (point A5) and 1222 m at 47.60 h (point B5), and
-
2.
for the 20-year design flood, the water level reaches 1214 m at 26.30 h (point A20) and 1222 m at 35.14 h (point B20).
As shown in the flood routing results, for both design floods (5- and 20-year), the dam is overtopped. Hence, it is necessary to assess the impact of the dam break flood on the downstream structures and villages.
4.3 Dam break flood at Hongshiyan
The dam break analysis at the Hongshiyan barrier lake follows the procedure described in Sect. 4.1. The discharge hydrographs for the 5- and 20-year design floods are shown in Fig. 15. Table 5 contains a summary of the dam break analysis results.
The analysis results show that for the 20-year flood, construction of the diversion channel can reduce the total released water from 3.54 × 108 to 3.15 × 108 m3. As will be shown in Sect. 4.5, this reduction in released water prevents the downstream Huangjiaoshu rockfill dam from being overtopped.
4.4 Flood analyses at Tianhuaban dam
The flood routing analyses through Tianhuaban dam were carried out by balancing the incoming dam break flood (Fig. 15) with the outflow through the discharge facilities according to the relationship as shown in Fig. 9. As explained in Sect. 4.1, the dam break flood at Hongshiyan can be used as the inflow to Tianhuaban. The calculation commenced at the initial water level 1040 m. Figures 16 and 17 show flood hydrographs at Tianhuaban for 5- and 20-year design floods, respectively. Table 6 contains a summary of the floods at Tianhuaban.
It is apparent that for all cases, the highest water levels exceed 1076.8 m which is the elevation of the dam crest. For the 20-year design flood, the depths of the overtopping flow at the dam crest are 4.55 and 4.16 m for with and without the diversion channel, respectively. On the other hand, it is generally known that a concrete arch dam can sustain overtopping for a period of time without being damaged. For example, the famous Vajont arch dam, one of the tallest dams in the world with the height of 261.6 m, survived with only small damages when a gigantic water wave overtopped it by as much as higher than 200 m (Kilburn and Petley 2003; Genevois and Ghirotti 2005; Alonso et al. 2010).
4.5 Flood analyses at Huangjiaoshu dam
Similar to the flood analyses through Tianhuaban dam, the flood analyses were carried out using the incoming floods from Figs. 16 and 17 with the hydraulic relationships in Figs. 10 and 11. The initial water level at Huangjiaoshu reservoir is controlled at 720 m. Figures 18 and 19 show the flood hydrographs at Huangjiaoshu for 5- and 20-year design floods. Table 7 contains a summary of the floods at Huangjiaoshu.
In order to ensure that Huangjiaoshu dam is safe, the maximum water level in the lake cannot exceed the elevation of the dam crest which is 775 m. Otherwise, the rockfill dam will break, resulting in another disaster. According to the results in Table 7, the only two cases in which the maximum water level elevation is lower than 775 m are the emergency work with the diversion channel. As such, the diversion channel is recommended as the emergency work.
While dam break flood analysis is not a precise science (Gee 2009), the results in Table 7, together with other results that are not presented in this paper, do provide a sound reason to adopt the alternative of having a diversion channel in order to minimize the catastrophic risk.
4.6 Diversion channel
As shown in Sect. 4.5, the construction of the diversion channel is a good solution to prevent possible catastrophes that may be triggered by the breaking of the Hongshiyan landslide dam. Historically, there are a number of landslide-dammed lakes that have used spillways or diversion channels to drain the water in the lakes, such as the Madison River Quake Lake in August 1959 in Montana, America (USACE 1960), the Mayunmarca rockslide dam occurred in April 1974 in Peru (Kojean and Hutchinson 1978), the Jiboa River landslide dam occurred in February 2001 in El Salvador (Bommer et al. 2002; Jibson et al. 2004), the Karli and Tung landslide dam in October 2005 in Pakistan (Schneider 2009), the Hunza rockslide-dammed lake in July 2010 in Kashmir (Kargel et al. 2010). In June 2008, an emergency channel (8 m wide and 13 m deep) was constructed at Tangjiashan landslide dam. This channel reduced the volume and the head of the released water during the dam breach, thereby created a controlled flood and prevented catastrophic consequences to 1.2 million residents downstream (Liu et al. 2010).
Based on the risk assessment of Hongshiyan landslide dam and the past emergency works at other sites, the diversion channel was decided to be built for the safety of the residents and structures in the downstream area, including Huangjiaoshu rockfill dam. After comparing the discharge and size of the channels and the construction time in other schemes, it was decided on a trapezoidal diversion channel with a side slope of 1:1.5 on both sides. The channel is 8 m deep with a base width of 5 m. The invert elevation of the channel inlet is 1214 m. The construction work started on August 8, 2014, and completed on August 12, 2014. Figure 20 shows the design ichnography of the diversion channel. Figure 21 shows the progress of the construction and the completed diversion channel.
4.7 Draining lake water through deep tunnel
In order to drain the water in the Hongshiyan barrier lake, a new spillway tunnel was built. This tunnel is located on the right side of Niulan River and is connected to the old diversion tunnel. This tunnel is also useful for rebuilding the power station. The invert elevation of the tunnel inlet is at 1095.5 m and the length of the tunnel is approximately 278 m. The cross section of the tunnel is rectangular with the size 7.5 m × 7.5 m. The discharge capacity of the tunnel is 840 m3/s when the water level in the lake is at 1176 m. Figure 22 shows the operations in the tunnel and the portal of the tunnel. On October 2, 2014, the spillway tunnel was completed and the water in the lake was successfully drained during the October 6, 2014, flood. Figure 23 shows the reappearance of the previously submerged Hongshiyan power station. In comparison, the submerged water line can be seen clearly.
4.8 Rebuilding Hongshiyan hydropower station
With the draining of the water in Hongshiyan barrier lake by the newly built spillway tunnel, the risk of dam break has been eliminated. However, the total water storage of the landslide dam is approximately 107 m3 with the backwater extending up to 24 km upstream. While the water in the lake poses a great threat to the residents living downstream, the cost of dismantling the landslide dam is huge. The demolition work is also extremely difficult as the rubbles of the dam body are large and there is no suitable location to stack them within the range of 30 km. It has been finally decided that the landslide debris should be upgraded to a permanent dam accommodating a power station with installed capacity of 200 MW. Figure 24 shows the design plan of the hydropower station. The projected is being constructed.
5 Concluding remarks
This paper documents various concerns involved in the decision makings for minimizing the hazards caused by a possible dam break flood induced by Hongshiyan barrier lake in a flood season. Highlights are as follows:
-
1.
According to the code “Standard for classification of risk grade of landslide lake (SL 450-2009),” the barrier lake is classified as “grade I extremely high risk,” for which a 20-year flood should be used for designing emergency works to prevent overtopping of the downstream Huangjiaoshu rockfill dam due to breaking of the landslide dam.
-
2.
The flood analysis indicates that the landslide dam would break in spite of the existing drainage tunnel provided by the submerged power station.
-
3.
An 8-m-deep diversion channel was constructed by the rescue soldiers worked in a difficult environment with the after-quake and falling rocks. The dam break and flood routing analysis indicates that this civil work can reduce the total amount of the released reservoir storage by 3.87 × 107 m3 and the peak dam break discharge to 8368 m3/s in a 20-year flood, ensuring that the Huangjiaoshu rockfill dam would not be overtopped.
-
4.
A large amount of dam break analytical work for different alternatives has been performed during the study. This was made possible due to our recent upgrading for dam break modeling and on the back analysis on a well-monitored case of the Tangjiashan barrier lake (Chen et al. 2015) and the computer program DB-IWHR 2014 that is coded in Microsoft Excel 2010, allowing fast, transparent and robust calculations for dam break flood predictions.
Response to the disasters caused by a barrier lake often involves considerations of many possible risks. Hence, it is important to have a key technology for performing the dam break flood analysis. The rescue actions described in this paper may be useful to other similar cases.
References
Alonso EE, Pinyol NM, Puzrin AM (2010) Catastrophic slide: Vaiont Landslide, Italy. Geomechanics of failures. Advanced topics. Springer, New York, pp 33–81
Baum RL, Crone AJ, Escobar D, Harp EL, Major JJ, Martinez M, Smith ME (2001) Assessment of landslide hazards resulting from the February 13, 2001. El Salvador earthquake. US Geological Survey Open-File Report, 01-119
Bishop AW (1955) The use of the slip circle in the stability analysis of slopes. J Geotech 5(1):7–17
Bommer JJ, Benito MB, Ciudad-Real M et al (2002) The El Salvador earthquakes of January and February 2001: context, characteristics and implications for seismic risk. Soil Dyn Earthq Eng 22(5):389–418
Chen Z, Ma L, Yu S, Chen S, Zhou X, Sun P, Li X (2015) Back analysis of the draining process of the Tangjiashan barrier lake. J Hydraul Eng 141(4):05014011
Dai F, Xu C, Yao X, Xu L, Tu X, Gong Q (2011) Spatial distribution of landslides triggered by the 2008 Ms 8.0 Wenchuan earthquake, China. J Asian Earth Sci 40(4):883–895
Evans SG, Delaney KB, Hermanns RL, Strom A, Scarascia-Mugnozza G (2011) The formation and behaviour of natural and artificial rockslide dams; implications for engineering performance and hazard management. In: Natural and artificial rockslide dams. Springer, Berlin, pp 1–75
Fan X, van Westen CJ, Xu Q, Gorum T, Dai F (2012) Analysis of landslide dams induced by the 2008 Wenchuan earthquake. J Asian Earth Sci 57:25–37
Fernández B, Salas JD (1999) Return period and risk of hydrologic events. I: mathematical formulation. J Hydrol Eng 4(4):297–307
Fread DL (1984) DAMBRK: the NWS dam break flood forecasting model. National Oceanic and Atmospheric Administration, National Weather Service, Silver Spring, MD
Fread DL (1988) BREACH: an erosion model for earthen dam failures (model description and user manual). National Oceanic and Atmospheric Administration, National Weather Service, Silver Spring, MD
Furumura T, Hayakawa T (2007) Anomalous propagation of long-period ground motions recorded in Tokyo during the 23 October 2004 Mw 6.6 Niigata-ken Chuetsu, Japan, Earthquake. Bull Seismol Soc Am 97(3):863–880
Gee D (2009) Comparison of dam breach parameter estimators. World environmental and water resources congress 2009, U.S. Army Corps of Engineers, Hydrologic Engineering Center, Davis, CA, pp 1–10
Genevois R, Ghirotti M (2005) The 1963 Vaiont landslide. G Geol Appl 1:41–52
Gorum T, Fan X, van Westen CJ, Huang R, Xu Q, Tang C, Wang G (2011) Distribution pattern of earthquake-induced landslides triggered by the 12 May 2008 Wenchuan earthquake. Geomorphology 133:152–167
Havnø K, Madsen MN, Dørge J (1995) MIKE 11-a generalized river modelling package. In: Singh VP (ed) Computer models of watershed hydrology. Water Resources Publications, Colorado, pp 733–782
Hydrologic Engineering Center (HEC) (2006a) HEC-HMS user’s manual. Version 3.1.0, CPD-74A, U.S. Army Corps of Engineers, Hydrologic Engineering Center, Davis, CA
Hydrologic Engineering Center (HEC) (2006b) HEC-RAS user’s manual. Version 4.0 beta, CPD-68, U.S. Army Corps of Engineers, Hydrologic Engineering Center, Davis, CA
Jibson RW, Crone AJ, Harp EL et al (2004) Landslides triggered by the 13 January and 13 February 2001 earthquakes in El Salvador. Geol Soc Am Spec Pap 375:69–88
Kargel JS, Leonard G, Crippen RE, Delaney KB, Evans SG, Schneider J (2010) Satellite monitoring of Pakistan’s rockslide-dammed lake Gojal. EOS 91:394–395
Keefer DK (1984) Landslides caused by earthquakes. Geol Soc Am Bull 95(4):406–421
Keefer DK (2002) Investigating landslides caused by earthquakes—a historical review. Surv Geophys 23(6):473–510
Khazai B, Sitar N (2003) Evaluation of factors controlling earthquake-induced landslides caused by Chi-Chi earthquake and comparison with the Northridge and Loma Prieta events. Eng Geol 71:79–95
Kilburn CRJ, Petley DN (2003) Forecasting giant, catastrophic slope collapse: lessons from Vajont, Northern Italy. Geomorphology 54(1):21–32
Kojean E, Hutchinson JN (1978) Mayunmarca rockslide and debris flow. In: Voight B (ed) Rockslides and avalanches, Peru, vol 1. Elsevier, Amsterdam, pp 315–361
Li MH, Hsu MH, Hsieh LS, Teng WH (2002) Inundation potentials analysis for Tsao-Ling landslide lake formed by Chi-Chi earthquake in Taiwan. Nat Hazards 25(3):289–303
Liu N, Zhang JX, Lin W, Cheng WY, Chen ZY (2009) Draining Tangjiashan Barrier Lake after Wenchuan Earthquake and the flood propagation after the dam break. Sci China Ser E Technol Sci 52(4):801–809
Liu N, Chen Z, Zhang J, Lin W, Chen W, Xu W (2010) Draining the Tangjiashan barrier lake. J Hydraul Eng 136(11):914–923
Ministry of Water Resources of the People’s Republic of China (2009) Standard for classification of risk grade of landslide lake (SL 450-2009)
Moussa R, Bocquillon C (1996) Criteria for the choice of flood-routing methods in natural channels. J Hydrol 186(1):1–30
Sato HP, Hasegawa H, Fujiwara S, Tobita M, Koarai M, Une H, Iwahashi J (2007) Interpretation of landslide distribution triggered by the 2005 Northern Pakistan earthquake using SPOT 5 imagery. Landslides 4:113–122
Schneider JF (2009) Seismically reactivated Hattian slide in Kashmir, Northern Pakistan. J Seismol 13(3):387–398
Schneider JF, Gruber FE, Mergili M (2013) Recent cases and geomorphic evidence of landslide-dammed lakes and related hazards in the mountains of Central Asia. Landslide science and practice. Springer, Berlin, pp 57–64
Singh VP, Scarlatos PD, Collins JG, Jourdan MR (1988) Breach erosion of earth-fill dams (BEED) model. Nat Hazards 1(2):161–180
U.S. Army Corps of Engineers (Omaha District) (1960) Madison River, Montana—report on flood emergency Madison river slide, vol 2. Omaha, NE
Wahl TL (2004) Uncertainty of predictions of embankment dam breach parameters. J Hydraul Eng 130(5):389–397
Zhou XB, Chen ZY, Li SY, Wang L (2015) On the standard of flood return period for high risk barrier lake in emergency disposal. J Hydraul Eng 46(4):405–413 (in Chinese)
Zhu YH, Visser PJ, Vrijling JK (2004) Review on embankment dam breach modeling. New developments in dam engineering. Taylor & Francis Group, London, pp 1189–1196
Acknowledgments
The authors appreciate the guidance and assistance provided by the Emergency Disposal Headquarters of the Hongshiyan Landslide Lake and Kunming Engineering Corporation Limited, Power China. The research work was supported by the National Basic Research Program of China Grant No. 2013CB036400 and National Natural Science Foundation of China (Grant Nos. 51309260, 51309190 and 51479168).
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Zhou, X., Chen, Z., Yu, S. et al. Risk analysis and emergency actions for Hongshiyan barrier lake. Nat Hazards 79, 1933–1959 (2015). https://doi.org/10.1007/s11069-015-1940-2
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11069-015-1940-2