(May 31, 1994)
SUGGESTIONS ON THE CONSTRUCTION SCHEME OF THE THREE GORGES PROJECT1
by Lu Qinkan2
According to the “Resolution on the Construction of the Three Gorges Project on the Yangtze River” passed at the March-April, 1992, fifth session of the seventh National People’s Congress (NPC), the NPC has resolved to approve inclusion of the construction of the Three Gorges project into the Ten-Year National Economic and Social Development Program (1991-2000).
The State Council shall, by taking into consideration the realistic conditions of the national economic development and the financial and material capabilities of the state, seek to organize implementation at an appropriate time; research shall continue towards the proper solution of the potential problems that have been identified.
This is a wise statement, which allows for the improvement of the Three Gorges project.
During the assessment of the project, I concerned myself mainly with the subject of flood control and was unable to take a serious look at the issue of sedimentation. Having recently investigated the available data in this field, I now realize that it warrants further attention.3 I hereby propose, in light of the NPC decision, that research continue towards the proper solution of the problems identified hereunder.
I. Sedimentation as an Obstruction to Navigation
The sedimentation experts’ research conducted during the leading group’s assessment has produced valuable data and increased understanding with regard to reservoir sedimentation. In order to reduce sediment aggregation in the proposed reservoir, the assessment report proposes an operational system that would “store the clear water and flush out the muddy.” This system, I believe, will be effective during normal times; that is, when there are no flood-control requirements for the middle and lower reaches of the Yangtze. However, flood control is one of the planned main functions of the Three Gorges project. During years of major floods, its reservoir must retain flood water, causing a tremendous amount of sediment to be deposited in the reservoir. In such situations the reservoir is unable to “store the clear water and flush out the muddy,” and has to “store the muddy water” instead. Sediment aggregation will occur in the back end of the reservoir, namely at the Chongqing harbor reach, thereby jeopardizing navigation.
At this point, it is necessary to refer to the findings of the “Experimental Research on the Impact of the 1954 Major Flood on the Chongqing Harbor Reach,” which the Institute of Water Conservancy and Hydro-electric Power Research (IWCHPR) released in July, 1988.4 If the Three Gorges reservoir were required to store flood waters of the magnitude of the 1954 flood5 and ensure the safety of downstream areas by maintaining a maximum allowable controlled flood discharge at Chenglingji,6 the water level in the reservoir would increase by six to eight meters in July, and seven to 13 meters in August. This would be a striking increase from normal flood years. The water level at the dam would increase from 145 meters (the flood-control level) to 175 meters, and the backwater level at Chongqing would increase from the normal flood level of 185.3 meters to 198.1 meters, causing a significant increase in sedimentation.
The IWCHPR study found that, at Jiulongpo port (the major upper port of Chongqing harbor) sediment deposition would concentrate along the wharf, causing the silted area to both increase in height and expand in area. The reservoir’s post-flood depth would be 185 to 186 meters, or more than 20 meters higher than the normal pool level. The original main channel would be obstructed totally, and the main flow would divert itself to the original supplementary channels. Flushing in the following two years would be unable to fully reopen the original main channel (it is still unknown how much more time would be required to do so), and the conditions for navigation would remain unfavorable.
The study also found that significant sediment aggregation would occur at Chaotianmen port (the main lower port at Chongqing harbor) at the mouth of the Jialing River. This sedimentation would cause the river’s main flow to change direction to Jinshaqi beach, where it would become rapid, shallow, and hardly navigable.
The IWCHPR report was released in July, 1988, and therefore was not considered in either the “Subject Report on Sedimentation Analysis” or the “Subject Report on Navigation Analysis” of the leading group’s assessment, which were released in February and April of the same year. Nor was the IWCHPR report considered by the ninth session of the leading group that examined the “Subject Report of General Planning and Water Level Analysis” in November, 1988, the tenth session from February to March, 1989, or the reporting assembly held by the state government in July, 1990. I wonder whether the IWCHPR report was even submitted to the meeting of the State Council Examination Committee of the Three Gorges Project in 1991?
The IWCHPR conducted a mathematical simulation of the major 1981 upstream flood.7 Although in 1981 the flow of flood waters in the middle and lower reaches was not large, had the Three Gorges dam been in place, the dam site water level would have risen from about 76 meters (the Gezhouba dam backwater level) to 147.25 meters (slightly higher than its flood-control level),8 and Chongqing’s flood level would have increased by 6.35 meters, to 199.05 meters. This level is higher than the 1954 flood backwater level of 198.1 meters, as described by the IWCHPR study. Compared with the physical simulation of the 1954 flood, the 1981 flood, though shorter in duration, had an even higher flood peak and a more concentrated sediment load; thus its threat to the Chongqing harbor reach was just as serious. These findings warrant a second physical simulation experiment.
Since the founding of the People’s Republic 40 years ago, the Yangtze River has experienced two different types of major floods: the first type is represented by the 1954 flood (downstream flooding), and the second by the 1981 flood (upstream flooding). We must be fully aware that either type can cause sediment buildup in the Chongqing harbor reach. In the case of a catastrophic flood on the scale of the 1870 flood,9 (which is to be the major flood-control target for the project), the reservoir would play an even greater role in flood storage, accommodating an even greater sediment load. This would lead Chongqing’s water level and level of sedimentation to grow to an even greater extent. Unfortunately, there has yet to be any research conducted on this issue.
During the assessment study, the sedimentation analysis was focused solely on making the concept of “store the clear water and flush out the muddy” operational. However, the analysis used only those years when the flood-control requirements were minimal and, therefore, when sedimentation of the reservoir was not serious. The analysis was preoccupied, in other words, with how, when small quantities of sediment accumulate at the Chongqing reach, optimizing reservoir management, undertaking harbor renovation, and improving navigation channels, could solve the problem. These efforts may have certain merits when the reservoir is filled only to the flood-control level. But once the Chongqing reach is turned into part of the Three Gorges reservoir and major floods occur, these efforts alone would be unlikely to serve as an effective solution.
The Ministry of Communication and the municipal government of Chongqing are pressing for the adoption of the higher pool level plan of 175 meters or 180 meters, so that Chongqing’s Jiulongpo port will be accessible to 10,000-tonne cargo ships. But the 175-meter pool level plan is likely to cause so much sedimentation in the Chongqing harbor reach after a major flood that the navigation to both the Jiulongpo and Chaotianmen ports would be obstructed. The 180-meter pool level plan would affect navigation even more seriously. If the main channel, once silted up, remained difficult for navigation even after two years of flushing, Chongqing would likely become a “dead port.” This would be an extremely serious situation, one which I am not sure the Ministry of Communication and Chongqing municipal authorities fully understand.
II. The Problem of the Two-Phase Storage Scenario
To reconcile the higher- and lower-pool level plans, a two-phase storage scenario has been proposed by the assessment, as stated in the “Subject Report of General Planning and Water Level Analysis.”
In the initial period, the normal pool level would be set at 156 meters so that the backwater level would reach only the mouth of Tongluo Gorge (about 16 kilometers below Chongqing), thus relieving Chongqing harbor and the mouth of the Jialing River from sedimentation and allowing sediment deposition to be more easily monitored.
The subject report noted that: “A period would be made available for the survey and investigation of reservoir sedimentation at the lower pool level.” However, this measure is aimed only at eventually raising the pool level to the proposed higher level.
The survey and investigation of the sedimentation situation must be conducted in all hydrological conditions (major floods in particular) over a period of a few decades. If the construction scheme allows for only a few years of study at the lower pool level, it will be impossible, I am afraid, to conduct investigations.
Based on the conclusion furnished by the Experts’ Group on Sedimentation, the “Subject Report on General Planning and Water Level Analysis” declares:
The sedimentation problem to be incurred by the lower pool level plan (150 meters to 160 meters in depth) is relatively easy to grasp, and its solution relatively reliable, whereas it is already evident that the higher pool level plan (170 meters to 180 meters in depth) will cause so much sedimentation as to affect, to varying degrees, the city of Chongqing, the harbor, and the mouth of the Jialing River.
This statement is both theoretically and empirically grounded. By proposing to conduct surveys and investigations during a lower pool level period, the two-phase storage scenario illustrates a lack of confidence in the higher pool level. If the higher pool level is not recognized as unattainable until after the completion of the dam, many of the facilities of the Three Gorges project will be inappropriate or uneconomical. Moreover, a great number of households, resettled according to the plan for “population relocation for development,”10 will also seek to return to their land.
Sediment deposits caused by the lower pool level plan and those caused by the higher pool level plan are different by nature. The former would be simple formations deposited in the river course, but the latter would be more complex formations deposited in the Chongqing harbor reach and at the mouth of the Jialing River. Even after a prolonged period of time, the survey and investigation of sedimentation caused by the lower pool level plan would still be inadequate for explaining the different and more threatening sediment deposits that will be caused by the higher pool level plan.
We should seek answers to these problems through scientific research and experimentation. Thus far, however, only one attempt has been made to simulate the 1954 flood. More flood simulation studies may be worthwhile. What is unacceptable is a gigantic field experiment on the Yangtze River. One trial experiment of this kind might well result in the suspension of navigation for several years. How can anyone risk such devastation?
The issue of navigation poses a dilemma: On the one hand, a lower pool level is preferable because it would cause less sedimentation. On the other hand, a higher pool level is preferable for enabling 10,000-tonne cargo ships to reach the Jiulongpo port facilities directly. If the 175-meter plan were subjected to scientific scrutiny, the results would show the extent of the sedimentation threat at Chongqing harbor, and those who now insist on a higher pool level would recognize it as unrealistic. Demands for increased transportation capacity can be realized in other ways-for instance, through the construction of reservoirs on the Yangtze’s upstream tributaries the depth of the river near Chongqing can be increased.
Another plus that those favoring a higher pool level often cite is the expectation of increased flood-control capacity. Important as this is, flood control should not compromise navigation to Chongqing harbor. Yangtze River floods can be enormous. Their control, therefore, cannot depend on only one facility (the Three Gorges dam), but must also be based on facilities built on the river’s upstream tributaries.
According to the “Main Points of the Yangtze River Comprehensive Development Program” (revised in 1988) by the Yangtze Valley Planning Office (YVPO), a number of reservoirs on upstream tributaries are slated for completion by 2015. At the same time, according to the “Report of Strategic Research on the Southwest China Hydro Power Energy Base” by the Hydro-electric Power Institutes of Sichuan, Guizhou and Yunnan provinces and organized by the Chinese Society of Hydro-electric Engineering, a number of large hydro power projects will also have been completed by 2020. These programs cover 70 percent of the drainage area of the Yangtze River Valley upstream of the Three Gorges, and include projects which, in terms of their total active storage capacity, will play a far greater role than the Three Gorges project in flood control both upstream and downstream.
III. The Submergence Area and Resettlement
1. Sedimentation. According to the “Regulations for Design to Deal With Reservoir Submergence of Hydraulic and Hydro Power Projects,” which the Ministry of Water Resources and Electric Power formulated in 1984,
in defining the backwater submergence zone for reservoirs on rivers bearing high sediment content, the designer shall take into consideration the impact of sedimentation over a period of time, normally 10 to 30 years from the time the project goes into operation, and pay attention to sedimentation at the upstream reaches of the reservoir.
But the study, by not adequately examining the numbers of people forced to resettle because of sedimentation, has grossly underestimated the number of people who will be affected. Moreover, the study examined only the impact of sedimentation on backwater in the main channel of the Yangtze, but not in its tributaries.
In January, 1988, the “Subject Report on Resettlement Analysis,” submitted to the leading group by the experts’ group on resettlement, stated:
The sedimentation-caused 20-year backwater level will not materialize until a few decades after the commencement of the project. With the construction of upstream reservoirs and improvements in soil conservation, the quantity of sediment discharge will change, possibly leading the backwater level to decline. Therefore, it is suggested herein that, for the time being, sedimentation-related resettlement be excluded from the resettlement investment budget during the assessment report.
In fact, the Three Gorges project will begin to cause sediment build-up in the reservoir in its 12th year, when the dam comes on line for power generation. By the completion of the resettlement plan, in the 20th year of the project, sediment will have been building up for eight years. Then, after only 12 years of the project’s normal operation, the impact of sedimentation will be felt fully. It is true that by that time, more reservoirs will have been constructed in the upper reaches of the Yangtze to meet the rising demand for power supply. However, they will control, at most, 70 percent of the drainage area, and leave the rest of it unattended. At the same time, any soil conservation projects would take much longer to manifest their effects. Therefore, enough sedimentation will build up in the near future to raise the backwater level of the Three Gorges reservoir. By not anticipating the sediment aggregation of the reservoir, the plans to construct reservoirs on the tributaries have actually violated the “Regulations for Design to Deal With Reservoir Submergence of Hydraulic and Hydro Power Projects.” They are inappropriate, and should be corrected in the subsequent process.
2. The Reservoir Backwater Curve.11 According to the Yangtze River Scientific Research Institute (YRSRI) calculation, if the 175-meter normal pool level plan is adopted, a 100-year flood after 100 years of sedimentation would raise Chongqing’s backwater level by 4.79 meters to 199.09 meters. Comparing this figure with the IWCHPR simulation, which indicates that after 109 years of sedimentation, a flood of the magnitude of the one in 1981 (equivalent to a 45-year flood) would be enough to raise the index by 6.35 meters to 199.05 meters, the YRSRI calculations appear to be conservative. The backwater curves of five-year, 20-year and 100-year floods, calculated by YRSRI, do not consider the corresponding sediment load carried by the flood, whereas the IWCHPR figures do. Every flood of the Yangtze carries a sediment load, and the greater the flood the more silt contained. Calculation of the flood backwater curve for a river with high sediment content, without consideration of its sedimentation impact, is not realistic, and cannot predict flood levels with any degree of safety. The backwater calculation of the Sanmenxia Gorge project, on the Yellow River, erred by using a “clear water” approach (not considering the sediment load of the flood). Its backwater calculation was that, based on the normal pool level of 350 to 360 meters above sea level, even a 1,000-year flood would not in any way affect Xi’an City. The fact is, however, that in the very first year of the project’s operation, when the pool level reached 333 meters (substantially lower than the proposed normal pool level of 350 to 360 meters), the safety of Xi’an was already under threat even though it had only been a normal flood year.
3. The Flood Submergence Standard for Chongqing. According to the Ministry of Water Resources and Electric Power’s “Regulations for Design to Deal With Reservoir Submergence of Hydraulic and Hydro Power Projects,” the submergence flood standard for important cities shall be set for a 50- to 100-year flood. The State Council has agreed that the flood-control standard for Chongqing should be set for a 100-year flood. Presently, however, the flood storage level indices furnished by the “Subject Report on Resettlement Analysis” have either excluded the 20-year sedimentation impact or have been derived from a “clear water” study for a 20-year flood, which falls far from the required standards of the above regulations. The results of the physical simulation and mathematical calculation of the 1954 and 1981 floods have already added 12.8 and 6.35 meters respectively to the natural flood level. These figures are much higher than is allowed by the above regulations. Also, based on these figures, the backwater level would be much higher than its present calculation, representing a substantial increase in the prospective submergence zone.
If the submergence zone of the project is defined according to the above regulations and by taking into account the above factors (such as sedimentation, reservoir backwater curve, and submergence flood standards), the resettlement population, which already exceeds one million, will increase by another several hundred thousand persons. This additional population will have an enormous effect on the resettlement plan, the natural environment and the investment budget.
What the foregoing discussion questions is whether the construction scheme for the Three Gorges project, namely of setting the normal pool level at 175 meters, of two-phase storage, and of population relocation for development, are ultimately feasible. I hope that these questions will be granted serious consideration.
By scaling down the normal pool level, the potential obstruction to navigation and the excessive resettlement requirements, the two most complicated problems facing the project, can be alleviated. This can, at the same time, contribute to reducing the budget, the intensity of the construction schedule, the complexity of technologies (such as the large shiplock cascade and the high shiplift) and the construction techniques required, environmental impacts, and safety hazards.
I hereby suggest that a comprehensive examination be given to the existing normal pool level plans. During the Three Gorges project assessment and the work of the State Council Examination Committee for the Three Gorges Project, many advisors and experts proposed, apart from the 175-meter plan, plans for 160-meter, 156-meter and even lower normal pool levels. Each plan needs to be tested in physical simulation analysis, for 1954, 1981, and 1870 floods assuming a 20-year accumulation of sediment. The same method should be applied to the study of five-year, 20-year, and 100-year floods with their respective sediment content, the sedimentation in the Chongqing harbor reach, and the prospective submergence zone. If any plan is adopted, the crest of the dam needs to be only five meters above the normal pool level. The plan for “over-storage” above the normal pool level (also the level of the resettlement zone) was strongly opposed in discussions during the assessment in every prefecture or county in the reservoir area. Taking into account the possible threat of sedimentation in Chongqing harbor, I would insist that the plans for over-storage should be officially dropped.
The 175-meter normal pool level plan was adopted by the leading group. Once the YVPO took it as the basis for its feasibility report, it was approved by the examination committee. Now, any attempt to restudy the subject must receive State Council approval. It is therefore suggested that the leading group and relevant experts meet to restudy this serious problem, then report their proposal to the examination committee and, upon the latter’s approval, submit it to the NPC.
Criticism or correction to any improprieties in the above opinions will be appreciated.
Sources and Further Commentary
1This letter, written on March 9, 1993, was not included in the original Chinese edition of Yangtze! Yangtze!
2Lu Qinkan is a member of the Economic Committee of the Chinese People’s Political Consultative Conference. He graduated with a degree in Civil Engineering from Zhejiang University in 1936 and obtained a master’s degree in hydrology in the United States. Since 1949, he has been working in the field of water resources and hydro-electric planning and was deputy chief engineer in the Ministry of Water Resources and Electric Power. In 1988, he was one of 10 experts who refused to sign the leading group’s assessment for the Three Gorges project.
3At the 1992 National People’s Congress meeting, Wang Demao, a delegate from Guizhou, saw a flaw in the proponents’ argument: proponents claim that the Three Gorges dam would protect downstream Dongting Lake from sediment aggregation, but also that sediment would be flushed through the dam leaving the reservoir available for storing flood water. Wang contented that the project cannot do both.
4The IWCHPR worked under the Ministry of Water Resources and Electric Power. This report publishes the experimental results from a hydraulic physical simulation model for the normal pool level of 175 meters.
5The 1954 flood devastated the middle and lower reaches of the Yangtze. 32,000 km2 of cultivated land were inundated, 19 million people were displaced and 30,000 were killed.
6Various sections of the Yangtze are also known by different names. Chenglingji is a control section on the middle reaches of the river, north of Dongting Lake.
7This large flood caused extensive damage above Yichang but none below it.
8The flood-control level is a low reservoir level proposed during flood seasons to provide storage for incoming floods. With a normal pool level of 175 meters the flood-control level of 145 meters leaves 221.50 x 106 m3 of flood storage capacity between the flood-control and normal pool levels.
9The peak flow of the 1870 flood at Yichang, 40 kilometers below the Three Gorges dam site, is estimated to have been 105,000 m3/s.
10This is a new concept, which came into being when the discussion of the Three Gorges project resumed in the early 1980s. The basic idea is that people relocated because of reservoir construction are no longer simply compensated, but are “rehabilitated,” with their lives and jobs taken care of under a unified and comprehensive plan to realize long-term stability.
11The water level in a reservoir will affect river height for a considerable distance upstream. The increase in height caused by the reservoir is referred to as the reservoir backwater curve.