environment international | beginning
of document | previous section
| next section |
CHAPTER 8: ASSESSMENT OF FUTURE-WITHOUT-PLAN SCENARIO
8.1 INTRODUCTIONA summary of the overall future-without-Plan (FWO) scenario is presented in Chapter 7 of the Regional Plan, Future Regional Development Context (pp. 77-97).
The information presented in this chapter reiterates and adds to the technical information, discussion, and scenario assessment for biophysical systems (hydrology, morphology, and biodiversity). Additional information beyond that presented in the Regional Plan is not available for social and institutional systems.
The information presented here pertains to (1) aspects of the FWO environment which will differ from the existing environment described in Chapter 4, and (2) additional information on earthquake and catastrophic flash flood hazards.
8.2 POSSIBLE RAINFALL AND FLOODING TRENDS AND IMPACTSFuture rainfall and rainfall variability cannot be predicted, not even as to whether they would increase or decrease. Observed trends over the past few decades are disturbingly large, however, and future increases or decreases could be similar in magnitude. Over the period 1901 to 1991, the shape of the regional rainfall pattern was remarkably stable, but annual rainfall increased moderately (10%), and its variability from year to year increased markedly (50%).(3)
Increases in rainfall and rainfall variability are consistent with predictions that monsoon circulations intensify with global warming (specifically, with increasing longitudinal temperature gradient). This is somewhat academic, however, as the current and future status of global warming is still a matter of substantial uncertainty, and its potential impacts on this particular region are unknown. Climate varies on all timescales for a variety of reasons and causality of phenomena can be difficult to establish.
Rainfall and flooding influence regional morphology, through their influence on sediment supply and runoff characteristics. The most sensitive subregions are the Meghalaya fans in the north and the Tripura piedmont streams in the south. The main lowland rivers such as the Upper Kushiyara (upstream of the Manu), the Upper Surma (upstream of Sylhet), and the Meghna will be less sensitive.
If rainfall and rainfall variability:
Stay at present levels, then conditions such as those experienced in the late 1980s and 1990s would prevail. Morphologic processes in the fans, piedmont streams, and elsewhere would continue at current rates.
8.3 TIPAIMUKH DAM/CACHAR PLAIN IRRIGATION PROJECT IN INDIA
8.3.1 Project DescriptionIndia has recognized the potential for constructing a major dam on the Barak River at Tipaimukh gorge for many years. In recent years a proposal has been advanced for a multi-purpose project that would provide hydro-power and flood control (see Regional Plan, Chapter 4, for project data). Information obtained through the Joint Rivers Commission provides a minimally adequate description of the project which has been used to make preliminary assessments of impacts on the region. Construction was proposed to start by 1993 but has been delayed pending resolution of various issues including the effects of flow regulation on Bangladesh. Regulation of the Barak's flow by Tipaimukh Dam would provide India with the opportunity to irrigate the Cachar Plain; this India proposes to do. Since this will involve a loss of water from the Barak, it is a matter of concern to the Northeast Region of Bangladesh. No statement is available as to how much water India proposes to take from this scheme. For the purposes of this study it has been assumed that the total depth of irrigation water to be applied is 1 m and that the water is diverted on a continuous basis during the six dry months (November through April).
Operational periodBased on this information it is clear that significant impacts on the region will result from implementation of the Tipaimukh Dam and Cachar Plain Irrigation scheme. During an average flow year these impacts would include:
Flood flows on the Barak River will be moderated, with peak flows at Amalshid being reduced by about 25% and flood water volumes being reduced by 20%. The corresponding water levels at Amalshid would be reduced by about 1.6 m. Similar changes would be expected along the Kushiyara River and upper Surma River. This should reduce the frequency of spills from the Kushiyara and Surma Rivers, reduces the extent of inundation in the Sylhet Basin and reduce channel erosion and sediment transport rates along the two rivers.
Dry season flows will be increased substantially (for example, average flows of the Barak River at Amalshid would be 4.2 times larger in February and overall dry season flow volumes would increase by 60%). This would increase water levels by 1.7 m at Amalshid. Increases in dry season water levels would also occur on the Kushiyara and Surma Rivers (for example, water levels during March should increase by 1.5 m at Sherpur). These increased dry season flows will provide benefits for navigation, irrigation, and fisheries, but could also reduce drainage from some areas.
These effects are documented further below in Section 8.7, Hydrologic Changes, which presents the results of the FWO regional surface water model runs.
Proposals for other dams on the Sonai and Dhaleswari Rivers (tributaries of the Barak) are not thought likely to be taken up before 2015. Therefore, no discussion of these proposals has been included in this study.
During reservoir filling (pre-operational period)Impacts experienced during the filling of the reservoir depend entirely on the operational rules adopted during the filling phase. Unusually low flow releases during reservoir filling can cause serious impacts on environmental systems in downstream areas; an example is the filling o the reservoir behind the Bennett Dam on the Peace River in Canada, which adversely affected ecosystems on the Peace-Athabasca delta. Filling as quickly as possible is done to maximize certain project benefits such as hydropower generation; slower filling represents a trade-off with other considerations such as downstream fisheries and farming.
Dam failureThis is documented in Section 8.8.2, Tipaimukh Dam Failure.
8.4 FWO STATUS, REGIONAL WATER RESOURCES INFRASTRUCTURE
8.4.1 Range of Possible Scenarios and Most Likely ScenarioThe FWO status of regional water resources infrastructure (embankments, water control structures, and river improvement works) is a matter for speculation. The range of possibilities is illustrated by the following sequence of alternative idealized future-without-plan scenarios, presented in order of declining water sector investment:
1. Frozen existing infrastructure + ongoing projects + plus business-as-usual new projects. Existing water control infrastructure stays in its 1993 condition; maintenance and repair just keep up with deterioration. Ongoing projects would be completed. These include Singar Beel and possibly extensive embankment raising in the Central Basin under the Systems Rehabilitation Project (SRP, described below). Beyond this, investment in implementation, operation, and maintenance of water control infrastructure continues much as it has in the past.
2. Frozen infrastructure + ongoing projects. Existing water control infrastructure stays in its 1993 condition; maintenance and repair just keeping up with deterioration. Ongoing projects would be completed. Beyond this, no new projects; no further rehabilitation.
3. Deteriorating infrastructure + ongoing projects. Maintenance is inadequate and as a result existing infrastructure deteriorates over time. Ongoing projects would be completed.
4. Frozen infrastructure. Existing water control infrastructure stays in its 1993 condition; maintenance and repair just keeping up with deterioration.
5. Deteriorating infrastructure. No new water control infrastructure is built. Maintenance is inadequate and as a result existing infrastructure deteriorates over time.
In real terms, either Scenario 2 or 3 is the most likely future scenario during the Plan.
8.4.2 Rehabilitation Activities Proposed Under Systems Rehabilitation ProjectRehabilitation of water resources infrastructure throughout the country is being addressed by SRP; this ongoing activity appears in Scenarios 1, 2, and 3 above. For the Northeast Region, SRP prepared feasibility studies of the rehabilitation of nine submersible embankment projects in the deeply flooded Central Basin (Table 8.1). Of these, five involve raising embankment design heights by over 1 m, over a total embankment length of 160 km. This is necessary because the gap between the pre-monsoon and monsoon flood levels has been decreasing. These five projects enclose a gross area of 27,000 ha, which is about 16% of the 172,000 ha within existing submersible embankment projects. If this approach is continued under subsequent phases of SRP, some or all of the other 14 submersible embankment projects in the region (gross area 90,375, embankment length 500 km) could be candidates for this type of upgrading as well.
NERP and SRP recognize that an increase in embankment design heights of 1 m or more will result in embankment crests (including the freeboard allowance above the expected pre-monsoon flood levels) at or above the 1:2 monsoon flood levels. The result will be increasing confinement of flows and sediment to the river channels, even if the areas are filled after the boro harvest. Post-monsoon drainage and fisheries could be adversely affected. For these reasons, SRP is recommending that rehabilitation of these projects be deferred and re-considered later, after implementation of the NERP initiatives Baulai River Improvement and Kalni-Kushiyara River Improvement, which are expected to decrease pre-monsoon flood levels significantly.
8.5 FWO CONDITIONS USED IN PROJECT PRE-FEASIBILITY AND REGIONAL SURFACE WATER MODEL STUDIESThe FWO conditions used in the project pre-feasibility studies are:
Historic rainfall and dischargesThe regional surface water model was run in two FWO configurations, one with and one without Tipaimukh Dam. The other prescribed FWO conditions were:
1991 water year
8.6 OVERALL MORPHOLOGIC CHANGESConstruction of flood control embankments, loop cuts, channel closures as well as ongoing channel changes over the last 20 years are responsible for a number of morphological adjustments that are currently underway in the region. These adjustments may take several decades to run their course, with impacts propagating long distances from the original point of disturbance. The most serious impacts will result from ongoing aggradation of the lower Kushiyara-Kalni River which is occurring as a result of several factors including upstream channel shifts, impacts of past loop cutting and alteration of the river's flow regime. Future developments would include:
The future sediment loads supplied to the region will depend primarily on future climatic conditions and the extent of land-use disturbances in the catchments. There is evidence from satellite photos that sediment yields from the Tripura Hills and Meghalaya Hills have increased in the last few decades. The main impacts from increased sediment yields would be reduction of land area suitable for agriculture, increased hazards to infrastructure and further reduction of fish habitat such as duars and beels. Increased sediment yields from Tripura will affect Piedmont streams such as the Juri, Manu, Dhalai, Karangi, Sutang, and Lungla Rivers. Increased sediment yields could accelerate ongoing sediment aggradation within flood control embankments on rivers such as the Khowai River and Chillikhali River. The overall affected area amounts to about 960 km2 or about 4% of the region.
Increased sediment yields from Meghalaya, on the northern boundary of the region, will increase channel shifting and sedimentation on the alluvial fans which extend from the Dauki-Piyain River in the east to the Someswari River in the west. The total affected area is about 1,400 km2 or about 6% of the region. Naturally occurring patterns of instability on alluvial fans will result in abandonment of some existing channels and development of new channels over time spans of ten to 20 years. For example, major avulsions appear to be either in progress or imminent on the Dauki-Piyain, Dhalai Gang, Jadukata and Someswari Rivers. However, channel avulsions are inherently unpredictable and could occur on any of the fans in the region over the Plan period.
In most cases, the impact of avulsions will be largely restricted to the fan areas. However, in the case of the Someswari River, an avulsion down the Atrakhali River would impact over much of the Kangsha River basin. For example, regional surface water model simulations indicated the avulsion would decrease discharges in the Kangsha River at Jaria Janjail by 250 m3 s-1, or 20 % of the monsoon peak and reduce flood levels on the Kangsha River by 0.3 to 0.5 m. This reduction would be offset by an increase in the eastward spills via the Atrakhali and Old Someswari Rivers. In other words, flood conditions will be reduced in one area but will be intensified in other areas.
A similar change is developing on the fan of the Jadukata River, where an avulsion is causing peak flows to spill westwards into Matian Haor and Tangua Haor. If trends continue, virtually all of the low flows will pass down the avulsion channel and Jadukata River will become essentially a dry channel most of the year. Important wetland habitat such as Tangua Haor will decrease in size as a result of the rapid sediment infilling that will occur after the channel shift is completed.
The net effects of such changes will result in both positive and negative societal impacts. Infrastructure, agriculture, and local residents near the newly developed channels will be seriously affected by increased flooding, erosion and overbank sediment deposition. Near the site of the abandoned channels, flood levels will be decreased while navigation will be impaired.
8.7 OVERALL HYDROLOGIC CHANGES
8.7.1 OverviewThis section summarizes the results of the FWO regional surface water model runs.
The cumulative effects of Tipaimukh Dam flow regulation, plus channel aggradation on the Kushiyara-Kalni River and Surma-Baulai River, would increase winter discharges and siltation along the Kalni River, and have the potential for raising the pre- and post-monsoon water levels by as much as 1.5 m at Markuli. During the monsoon season, however, the effects of channel aggradation and Tipaimukh flow regulation largely offset each other, so that the peak water levels were increased by about 0.3 m. The result of all this would be greater depth and extent of flooding during the monsoon season, retarded drainage during the post-monsoon season and earlier and more severe pre-monsoon flooding of unprotected areas adjacent to the river.
Changes along the Surma-Baulai River were found to be similar in nature but smaller in magnitude than on the Kushiyara-Kalni. For example, at Sukdevpur on the Baulai River, water levels during the monsoon season were found to be virtually unchanged from existing conditions. During the post-monsoon season, water levels were raised by about 0.8 m. At Bhairab Bazar, on the Meghna River, water levels and discharges were found to be only slightly affected by upstream changes in a year similar to 1991, but may be more significant during drier years.
8.7.2 Detailed Water Levels and Discharges by River SystemThe information presented in this section is based on the output of the Northeast Regional Model simulations based on the 1991 water year (see Section 7.2.1). The changes noted are relative to current conditions.
Surma-Kushiyara systemThe Tipaimukh Dam/Cachar Plain Project on the Barak in India will substantially alter discharges of the Barak where it enters Bangladesh at Amalshid. Available information suggests that monsoon peak flows would decrease by about 30% (from 5250 to 3500 m3 s-1). Winter flows would double or triple, increasing 100 to 200% (from between 170 and 250 to 500 m3 s-1). Of the monsoon peak decrease, monsoon flow in the Surma and Kushiyara Rivers would decrease by 800 m3 s-1. Surma-Kushiyara and Surma-Sarigoyain floodplain discharges would decrease by the remainder of 1150 m3 s-1.
The Surma and Kushiyara Rivers along their entire lengths, and part of their tributaries, are also affected. At Fenchuganj on the Kushiyara, for example, model monsoon peak flows decreased by about 20% (from 2900 to 2400 m3 s-1). Peak levels decreased by 1 m (Regional Plan Figure 21A). Model winter flows increased by about 80% (from 250 to 450 m3 s-1). Levels increased by almost 2 m.
Further downstream in the Kalni-Kushiyara, model water levels increased by as much as 0.3 m in the monsoon and 1.5 m in the winter and pre-monsoon periods, as a result of sediment deposition. The affected reach extends as far as Ajmiriganj. By Bhairab Bazar, model flows and levels are almost unchanged from current conditions. In a simulation based on a drier year than 1991, however, model winter flows might increase significantly; simulated discharge hydrographs show that Bhairab Bazar winter flows are highly variable due to tidal effects.
Similar but somewhat smaller changes occur on the Surma (Regional Plan Figure 21B). At Kanaighat, model monsoon levels decrease by 0.5 m, while winter levels increase by 1 m or more. At Sukdevpur model water levels are almost unchanged from current conditions.
Someswari-Kangsha systemSomeswari-Shibganjdhala siltation and growth of the Atrakhali, a new distributary of the Kangsha, will reduce Kangsha discharges along a 50 km reach from Sarchapur and Mohanganj, centred on Jaria Janjail (Regional Plan Figure 21B). At Jaria Janjail, model monsoon water levels decreased by 0.3 to 0.5 m, and monsoon peak flows decreased by 20% (from 1250 to 1000 m3 s-1). Discharges in the Atrakhali and the Old Someswari channels increased by similar amounts, reflecting the diversion of flow away from the Kangsha and into the Atrakhali-Old Someswari.
Jadukata RiverAs the ongoing avulsion into the Patnaigang continues, Jadukata non-peak flows will eventually cease and it will become essentially a dry channel most of the time. Two-thirds of peak discharges would flow along the Patnaigang into Matian Haor, and the remainder would flow along the Jadukata.
Khowai RiverThe model incorporates estimated future aggradation in the reach from Shaistaganj downstream to Habiganj. As a result of this, model water levels increase 1 to 2 m near Habiganj and slightly near Shaistaganj. These higher levels imply embankment overtopping and greater risks of breaches in this reach. If the aggradation were extended further up/down the river, the higher levels would extend with them.
Changes in other boundary riversOnly localized changes in other boundary rivers are expected, therefore these were not incorporated in the model.
Summary and conclusionsThe most significant changes are those associated with:
The potential Tipaimukh Dam/Cachar Plain Project, which the model indicates would decrease upper Surma and upper Kushiyara monsoon peak levels by 1.5 m and increase winter discharges by 100 to 200%, for conditions similar to those in 1991.
8.8 LIKELIHOOD AND POSSIBLE IMPACTS OF HAZARDS
8.8.1 EarthquakesThe region is known to be vulnerable to earthquakes. These events, though relatively rare are extreme in intensity, and can reverse existing morphologic trends and even induce re-configuration of the drainage system. The likelihood that during 1991-2015 the region would experience an earthquake of magnitude 7.6 (similar to the 1918 event, return period of 30 to 50 years) is between 40 and 60%; of magnitude 8.7 (similar to the 1897 event, the largest on record, return period of 300 to 1000 years) is perhaps 2 to 5%, assuming the events are random and can be described with a simple binomial probability model.
On past evidence, river channels and sedimentation patterns in the Northeast Region may be subject to major disruptions following a severe seismic event. During past earthquakes, instances of ground liquefaction, landsliding, rapid subsidence, collapse of river banks, and changes to river courses have been documented (District Gazetteer, 1917). The effects of earthquakes along the Brahmaputra River were described by Oldham (1899), reproduced in Chapter 4 of French Engineering Consultants (1989):
Strong ground shaking triggers liquefaction of river cross-sections in a few seconds; underwater slopes slide towards the stream axis, the bottom of the river heaves, and the banks become lowered;
Water immediately starts to rise and overflows the banks and adjacent zones where infilling of the channels takes place. Natural sills form, causing temporary lakes to develop;
Channels gradually re-open by scouring where currents are strong enough, and consequently water levels decrease. Where channels remain blocked, streams desert their old channels to form new ones; and
In subsequent years, the huge amounts of sediment poured into the river as a result of the earthquake gradually move downstream. Sediment transport is higher than previously and siltation conditions are therefore modified.
Earthquakes are believed to have also induced landsliding and slope failures in headwater catchments in the Shillong Plateau, which could greatly increase the amount of sediment supplied to the region for long periods of time.
Joglekar (1971) described apparent impacts of major earthquakes on the upper Brahmaputra in Assam, India. After the severe earthquakes of 1947 and 1950, the bed level near Dibrugarh rose substantially. Between 1947 and 1951, low water levels rose by as much as three to four metres; thereafter they were steady.
8.8.2 Tipaimukh Dam FailureAssessment of this type of risk and risk management planning should be undertaken by a qualified dam safety specialist, and we are uneasy as hydraulic and environmental specialists making comments on this subject. This risk is however a significant issue relating to future environmental management of the Northeast Region water system which we feel a responsibility to mention here.
A dambreak is a catastrophic failure of a dam which results in the sudden draining of the reservoir and a severe flood wave that causes destruction and in many cases death downstream. While such failures are rare and are not planned they have happened to dams, large and small, from time to time. The International Commission on Large Dams (ICOLD) has identified 164 major dam failures in the period from 1900 to 1965 (Jansen, 1983).
With respect to the safety of the proposed Tipaimukh Dam, it is our impression that well-designed and constructed rockfill dams are perhaps the safest type for large heights (Tipaimukh would be among the largest of such dams in the world; see rockfill dams listed by Cooke, 1984), but we imagine that local circumstances may be much more important in this respect than dam type.
Two examples illustrate the types of failures that have been reported. The most famous, the Teton Dam in the United States, was a 90 m high earth-fill dam which failed in 1.25 hours. The flood wave which was released had a peak discharge of 65,000 m3 s-1 at the dam and a height of 20 m high in the downstream canyon. The Huaccoto Dam in Peru was 170 m high, similar to the Tipaimukh Dam; it failed over 48 hours due to a natural landslide in the reservoir.
Generally, a flood wave travels downstream at a rate in the order of 10 km hr-1 although velocities as high as 30 km hr-1 have been reported near failure sites. From these wave velocities, it would appear that the initial flood wave could travel the 200 km distance from Tipaimukh Dam site to the eastern limit of Bangladesh within 24 hours having a height of perhaps 5 m. Peak flooding would occur some 24 to 48 hours later. High inflows would persist for ten days or longer and the flooded area would likely take several weeks to drain.
The Tipaimukh reservoir is huge (15,000 Mm3) compared with experience reported in the literature. In the event of a significant unplanned discharge, the river system in Bangladesh would respond (drain) rather slowly, as characterized by the outflow rate relative to the floodplain storage volume), such that most of the water released would remain ponded over the Northeast Region for some time. Assuming a release volume of 10 Mm3 and a ponded area of 100 km2, the depth of flooding would be an average of 1.0 m above the normal flood level.
There will be a need for Bangladesh and India to cooperate in formulating and implementing risk management measures. A wide range of risk management measures are normally undertaken (Jansen, 1983), including: regular inspections by independent engineering teams, instrumentation and plans for warning downstream populations of deteriorating conditions of a dam, evacuation plans, and so on. As and when India's plans proceed, there will be a clear need for Bangladesh to avail itself of expert technical assistance from dam safety specialists experienced with very large dam/reservoir systems and trans-border risk management.
For illustrative purposes only, we show modelled flood waves for a test case of a instantaneous failure, 50 m wide extending to 100 m below the crest of the dam. Discharge and water level hydrographs are presented for three locations (Figure 11): at the exit from the mountain valley (km 80), at Silchar (in the middle of the Cachar plain, km 140) and at Amalshid (km 200). It can be seen from this that substantial attenuation of the flood wave would occur upstream of Amalshid and that the flood wave at Amalshid is a long-duration event. Depending on the breech geometry and peak discharge, the flood peak would occur at Amalshid approximately 2 to 3 days after the dam break had occurred and flooding would continue for ten days or more. The flood levels at Amalshid would rise to approximately 25 m PWD, which is at approximately 8 m above the floodplain level. This flood level depends on the boundary assumptions which were made and could be less depending on floodplain conveyance.
8.8.3 Catastrophic Flash Flood Events in the Meghalaya Alluvial Fans and Tripura PiedmontFive river reaches in the Meghalaya border area and four in the Tripura border area were identified as vulnerable to catastrophic flash flood events. Each channel is a nodal reach, lying between an upstream basin fed by dendritic tributaries, and a downstream aerial delta. Interviews with inhabitants along these reaches and theoretical studies strongly suggest that flash flood events in these reaches take the form of bores or walls of water, which generate forces sufficient to capsize boats and to sweep away people in or near the river channel. These catastrophic impacts are in addition to flood damages to infrastructure, crops, and so on, which also occur when water levels rise more gradually.
The annualized death rate from catastrophic flash flooding, totalled over the 2600 villages along the nine vulnerable river reaches (total population 2 million) could be on the order of 100 to 1500 persons per year (Figure 12). This estimate is based on reports of 49 flash flood fatalities in 1988 in five villages on the Meghalaya rivers (of which 45 were from a single boat capsize accident), and four fatalities in 1993 from four villages on the Tripura rivers (Improved Flood Warning, NERP concept paper). The implied death rates are 15 per thousand and 1 per thousand for these two events respectively, using regional average figures for households per village (130) and persons per household (5.4). The corresponding annualized death rates are 100 to 1500 persons per year, if return periods for these death rates of 1:100 and 1:10 years respectively are assumed. An increasing rainfall trend could lead to a dramatic increase in death rates.
| next section |
This web page (previously international environment & development professional's home page) created by water environment international. All site pages (c) wei. Last modified 26 Aug 2002 . Comments? Problems? Email email@example.com
You are visitor since 25 Aug 98