Dams and Reservoirs

Overview

After analysis of the basin, only three big dams / reservoirs appear to have a strong influence on the flood events on Maritza and Tundja.

• Topolnitza reservoir operated also by “Irrigation Systems” PLC;
• Vacha dam operated by Bulgaria’s National Electric company;
• Zrebchevo reservoir operated by “Irrigation Systems” PLC.

These dams with large retention capacity (above 100 Mm3) receive direct inflow from rivers of Maritza and Tundja basin with significant upstream watershed (above 500 km²).

This does not mean that the other reservoirs do not influence the flood regime of the watershed, as we will see later. But at this stage we meant that only these three big Dams and reservoirs may be overpassed by large floods (such as Topolnitza reservoir in 2005) and thus provide a strong discontinuity effect on the flood generation.

Other dams or reservoirs within the watershed have large capacities but gather waters form much smaller sub-basins (or accumulate waters from other watersheds through artificial channels or pipes) and their effect during rather large floods may be considered as secondary when compared with these three “big” ones.

The statistics of river discharges after the settlement of these dams have been largely modified and it is obvious that Dams have a positive impact for flood protection at least until a certain level of flood intensity. The flood of August 2005 (cyclone Ihtiman) where Topolnitza dam was overtopped which led to big damages downstream, should remind us this fact.

For flood management we are interested mainly in two aspects of these dams and reservoirs.

• What is the probability of Dams overflowing in case of very strong events (which oblige us to know their water security level, the weir crest, the retention curves, the number of gates...): This analysis in necessary to define the flood scenarios to be used for the production of flood maps.
• What are the operation rules during large floods : we need to know and make a diagram of these rules in order to incorporate them in our hydrodynamic models.

1. The Topolnitsa reservoir

The Topolnitza reservoir was constructed in 1963. It gathers waters from Topolnitza and Mativir rivers in the mountainous upper part (North western) of Maritza basin about 60 km upstream from Pazardjik.

The upstream watershed or drained area is about 1381 km²
 
The reservoir capacity (at security level) is 141 Mm³
 
The lake surface at this volume is 9.6 km². The retention curves are available.
 
With a height of 78m, it has four removable gates of 5 m height and 4m width with equal capacities. Once a gate is removed with water above the weir crest it cannot be reset.
 
Its spillway is designed for a 1000 m³/s discharge corresponding to a 0.1% yearly probability flood (1000 years). The peak inflow discharge for a 1000 year flood has been estimated at 1500 m³/s.
 
The security rules for flood management have set recently – after August 2005 - to 45 Mm³ the reserve capacity below the crest.
 
The Topolnitza river downstream is calibrated for a maximum 330 m3/s discharge
 
The reservoirs allows for a 60 m³/s bottom release through two pipes of 1,4m diameter each;
 
Beside, an electricity plant can turbine up to 30 m³/s

Several data were obtained in order to model and analyse the reservoir operations during floods:

• Retention curves (Volume as function of water elevation);
• Daily inflows and outflows from 2001 to 2006;
• Hourly outflows and inflows during the 2005 event;
• Outflow law for claps (Q = f(h)).

2. The Vacha Dam

The Vacha dam is an electrical dam on the Vacha river of 226 Mm3 Capacity constructed in 1975.

It is followed in cascade by the Kritchim dam which is of much lower capacity (20 Mm3) and which would be rapidly overpassed if the Vacha dam was itself overwhelmed. Kritchim dam may thus be considered as transparent for flood propagation in the Vacha cascade.

The upstream watershed of Vacha dam in the Rhodopes mountains is 1465 km²
 
The lake surface at this volume is ??? km².
 
Vacha dam (about 110m height) is equipped with four removable gates 8m wide and 3m high (weir crest : 528,8 m ; Valves crest 535,8 m).
 
The Dam crest is at 540m altitude with a probability of over toping of 0,01 % (10 000 years) corresponding to a flood with peak discharge of 2720 m³/s, the spillway itself being designed for a 2060 m³/s discharge for this same project flood.
 
The security rules have set recently to 28 Mm³ the reserve capacity bellow the crest for flood management.
 
The Vacha river downstream is calibrated for a maximum 300 m³/s discharge.
 
The reservoirs allows for a 160 m³/s bottom release through two pipes of 2,2 m diameter each;
 
Beside, the electricity plant can turbine up to 50 m³/s

No data was obtained on this Dam which was excluded of the scope of this project. The Vacha river study and integration in the modelling system will be part of a follow-on project.

3. The Jrebchevo Reservoir

The Jrebchevo reservoir was constructed in 1973. It is a mixed irrigation / electrical dam on the Tundja river of 400 Mm3 capacity.

The upstream watershed of Jrebchevo dam in the Balkan mountains is about 800 km2.. It is located just upstream the city of Banya and downstream another big dam (Koprinka dam, 140 Mm3, which itself intercepts waters flowing from the Balkan mountains for another 600 sq km)
 
The lake surface at this volume is about 24,9 sqkm. The retention curves are available.
 
The Jrebchevo reservoir (about 50m height) is equipped with 5 claps, which are manipulated by chains. Each clap dimension is 10,5 m long and 5 m high. The claps are moved by hand or by electric machines depending of the water level.
 
The Dam crest is at 269,5 m altitude with a probability of over toping of 0,1 % (1 000 years) corresponding to a flood with peak discharge of 1630 m3/s, the spillway itself being designed for a 1400 m3/s discharge for this same project flood.
 
The security rules have set recently to 100 Mm3 the reserve capacity bellow the crest for flood management.
 
The Tundja river downstream is calibrated for a maximum 250 m3/s discharge at least from Banya to TSavoy
 
The reservoir allows for a 150 m3/s and another 81,2m3/s bottom release through two pipes of 3,5m and 3.2m diameter equipped with claps;
 
Some more 30m3/s release may also be diverted toward Maritza basin through pipes.
 
Beside, the electricity plant can turbine up to 41 m3/s

As for Topolnitza reservoir various data were obtained for analysis:

• Retention curves (Volume as function of water elevation)
• Daily inflows and outflows from 2001 to 2006
• Outflow law for claps (Q = f(h))

Analysis performed

Only the Topolnitza reservoir was modelled in the project.

Vacha Dam and Jrebchevo reservoir are both out of the scope of the modelled system.

However because of the necessity to run flood scenarios we needed to estimate in which conditions the Jrebchevo reservoir could be overtopped with a 100 Mm3 security volume.

Beside, the 2005 event data were used for calibration of the reservoir model of Topolnitza within the global hydrological and hydraulic model of the Maritza watershed.

In order to asses the possibility of dam overtopping accounting for the defined “security volumes”, the daily inflows have been used to perform a statistical analysis on the probability of high volume inflows for periods of one to 5 days.

Actually with only 6 years of data such an analysis cannot be considered as very sound. The idea was also to compare the probabilities computed, with the values which were used as project values for the computation of the weirs and gates.

We provide hereafter the results of this analysis (adjustment of extreme values on a Log–Log law):

On these graphs we reported the computed return values of inflows which were obtained by the engineers during the Dams designing project in order to compare their estimations with ours.

Jrebchevo DAM
Statistics on 1 day inflows	Statistics on 3 Days inflows
Topolnitza Reservoir
Statistics on 1 day inflows	Statistics on 3 Days inflows

Note that for Zrebchevo Dam, our analysis made upon only 6 years of data is totally relevant with the statistics elaborated during the design project of the reservoir.

For Topolnitza, where we have the 2005 flood taken into account in our statistics analysis, we can see that the design values are positionned largely underneath the statistical trend that may be chosen for adjustment of our sample;

However, as the design data show, the event of 2005 would correspond to a 100 year type event, which is in our opinion close to the reality. Consequently we will also consider that the design statistical laws are relevant for Topolnitza.

For Zrebchevo dam, we may conclude from this analysis, that a 100Mm3 security volume prevents the overtopping of the claps for events of return period below 500 years. For 1000 years return period floods the Reservoir would open its gate after one day which would provide downstream discharges of about 1000 m3/s for several hours.

Consequently, the “1000 year scenario” on upper Tundja will have to be done with a 1000 m3/s discharge.

TheTopolnitza dam was already overtopped in the past. The adopted 45 Mm3 security volume corresponds to an inflow event of about 25 years return period. The full opening of the claps would not occur before 100 year return period events.

In 2005, the flood occurred when the reservoir was full after several flood events. Consequently they had to open the four claps of the reservoir which lead to more than 600 m3/s discharges at the peak of the flood.

Conclusions about Dams and reservoirs

In 2005 Topolnitza was overtopped by a severe flood of say, 100 year return period.

Actually, the reservoir was full when the flood came as the winter, spring and the beginning of summer had been very wet.

But, even if it had not been full, we are quite sure that the reservoir would also have been overtopped, while less severely and with minor damages downstream. The operation rules during floods, which could be somewhat upgraded, could only slightly reduce the damages due to Dam overtopping.

This event has to remind us that the reservoirs and dams do not protect the flood plains downstream for extreme events but only for the more frequent ones.

From our analysis based upon 7 years of data and the design values taken for these dams we may say:

Topolnitza reservoir with 45 Mm3 of security volume gives a total protection for flood under 25 years, a good protection (by reducing their effect) for flood between 25 to 75 years of return period and will be nearly transparent (offering some time lag however for alarm) to extreme events over 100 years of return period.

The Jrebchevo reservoir is the biggest of the three dams, with the smallest catchment area, and because of its surface, the one with the most retention and security volume. Obviously the Jrebchevo dam provides a full protection of the downstream area even for very extreme floods but until which extent?

From our analysis based upon 7 years of data and corroborated by the project floods used to design the dam structure, we can consider that floods over 1/500 years probability would start overpass the dam and that 1/1000 years flood would deliver more than 1000 m3/s into the Tundja river, which of course would not handle it.

The Vacha Dam with its 25 Mm3 of security and its 1465 sqkm basin in the Rhodopa mountains will provide security only for 10 to 20 years return floods. In 2005, two things protected the population downstream from having it overtopped : i) the rains in the Vacha watershed were less intense than in Western Rila mountains and ii) the Vacha waterlevel was well below its security level when the event occurred.

However we have several events in our records where we can see that Vatcha dam had to open its claps due to water rises in the retention lake during localised floods. The proximity of the Vacha river confluence with Plovdiv should suggest to take some specific warning measures for this river and its dams. Note that the design flood used for Vacha was more than 2700 m3/s (10-4 probability) which gives obviously a 100 year return period flood over 1000 m3/s. of which only a few part would be dampened.