Critical review of engineering aspects of Bhimkund reservoir and

Upper Lake constructed by Raja Bhoj.

K.G.Vyas*

Background

Raja Bhoj (1010-55) constructed Bhimkund reservoir (a very large pool of water) on
river Betwa, near Bhojpur (23º 6´ N and 77º 38´ E) – a small village nearly 30 km south
of Bhopal and Upper Lake on river Kolansh (tributary of Betwa) at Kamla park, Bhopal,
Madhya Pradesh, India. Many traditional stories exist amongst local people about
construction of Bhimkund. According to one, Raja Bhoj was stricken with serious illness
(perhaps leprosy) and could not be cured by his physicians. Ultimately, a monk was
approached and he, after examinations, pronounced that Bhoj would die of the disease,
unless he constructs a lake which is fed by 365 springs and streams. After taking bath in
this lake on a particular day and time, he would be cleansed. Stories tell that after long
and weary investigations, skilled engineers of king discovered the valley in Vindhyan hill
ranges which contained the head waters of 359 springs and streams. It was Betwa river
valley which was located east of his state capital (Dhar). The number deficit of streams
and springs was resolved by Kalia – a local Gond chief. He pointed out a missing river,
which along with its tributaries, made up the desired number. The engineers diverted the
missing river through a feeder lake (Upper Lake of Bhopal) to meet the desired number.
This missing river is named after Kalia Gond and is known as river Kaliasot.

W. Kincaid (1888) provided first brief account of Bhimkund in Rambles among ruins in
Central India. C. E. Luard (1908) also provided brief account of Bhimkund in Bhopal
State Gazetteer, but the same is mainly based on the findings of W. Kincaid.
Archeological Survey of India (ASI), many scholars and historians have also made
valuable contributions to enrich our understanding about the past history of central India
but these contributions lack in providing the engineering ability, skills and knowhow of
ancient people. This paper attempts to fill some gap and paves way for improving our
understanding about long lived perennial water bodies.

W. Kincaid described Bhimkund reservoir as the largest and the most beautiful sheet of
fresh water in peninsular India. According to his report, two earthen bunds (87 ft and 40
ft high) were constructed – one at Betwa and other to divert river Kaliasot at right angles
to meet Betwa. The pitching of these earth filled bunds was done with dressed blocks (4
ft x 3 x 2.5 ft) of local red sandstone. These blocks were laid on either side of the bund,
one on the other, without mortar, but fitting so truly as to be watertight. Kincaid says that
waste-weir, cut in solid sandstone, was located at the blunt apex of the triangular valley
and is nearly 3.2 km east of small bund. The report says that maximum water level
(MWL) of Bhimkund reservoir was 6 ft below top bund level (TBL).

Kincaid appreciates positioning of waste-weir of this grate reservoir saying that any error
in deciding the level would have washed away the dam. He describes it as a proof of
practical ability of Hindu engineers of tenth century. According to him, the area of
Bhimkund was nearly 250 sq miles (650 sq km or 65000 hectares) and the maximum
depth was nearly 100 feet. The dam wall is more than one kilometer long and now serves
as road connecting Mendua village with Bhojpur. It is shown below in Figure 1.

Figure 1, Dam wall between Bhojpur and Mendua village

Bhimkund bund (Figure 2) was broken by Hoshangshah in the year 1430. Broken bund is

Figure 2, Broken stretch of the dam wall near Shiv temple, Bhojpur

located near Shiv Temple, Bhojpur. Many stories (local as well as by contemporary
historians) exist to tell the reasons for destroying the bund by Hoshangshah. It is said that
local tribal army was deployed for destroying the bund across river Betwa. Gond army
took three months to destroy the bund and Bhimkund Lake emptied in nearly three years.

2
Cultivation started on lake-bed after nearly 30 years. Destruction of dam provided vast
fertile land for cultivation.

Traditions tell and Kincaid confirms that run-off contribution of Betwa river system was
insufficient to fill the Bhimkund reservoir so they skillfully met the deficiency by turning
the waters of another valley in Bhimkund. Kincaid says that diversion increased the
drainage area of Bhimkund by nearly 500 sq miles.

Critical review of civil engineering aspects of Bhimkund reservoir and

Upper Lake - Author’s Contribution

Bhimkund reservoir

Author studied various technical aspects of Bhimkund and Upper Lake to understand the
appropriateness and technical competence of tenth century engineering and has compared
it with modern civil engineering criterion used in dam designing, construction etc. The
study include aspects like site selection, bund construction, average and maximum water
depth, character of catchments, area (original and enhanced), run-off contributions,
storage capacity, shape and size of the reservoir, silt deposition, basin water diversion in
context of river science.

A. Site selection, bund construction and depth of Bhimkund.

Site selection – Author has reviewed the technical details available for Bhimkund dam
vis-à-vis present day technical criterion for site selection of dams. The review confirms
that technically sound and economically viable site (narrowest gorge) was chosen by
engineers of tenth century. It is, therefore said, that the site selection was perfect as per
present day technical and financial norms.

Bund construction – The bunds of Bhimkund are earthen bunds of unusual size. They
have been constructed by filling the compacted local soil and stones. The stone pitching
was done without mortar on side slopes so as to protect the bund from water action (wave
and rain water action). The workmanship was extremely good and the stone pitching of
dressed stone blocks made it almost water tight therefore the compacted soil did not
permit water entry / movement inside the bund. Bund, due to, use of durable and weather
resistant construction material has survived for more than 1000 years. The height of bund
was decided with such perfection that breach (due to excessive flood) did not occur in its
life time i.e. nearly in 400 years. The un-breached dam wall, even after witnessing
weathering impact for more than 1000 years, is still in good condition.

Depth – The basement or bed rock is basalt and sandstone. The assessment of its depth
was done from shorelines to main bund by using data of bore-hole logs at 18 different
locations. The bore-hole logs provided depth of basement rock and silt depth at different
locations.

3
The difference between reduced level (RL) of waste – weir and ground level (GL) of bed
rock provided the average water column which suggests maximum depth of water in the
reservoir. The maximum depth, at the time of construction, could be close to 40 meters.

Attempt was made to assess the average depth of the reservoir. Geomorphology and slope
of the reservoir area suggests that its average depth could be nearly 20 meters.

Catchments character and its size (original and enhanced)

Catchments character – Catchments of Betwa river-system were located in Vindhyan hill

ranges situated in north-west, west and south of reservoir. The delineated gross catchment
is shown in the digitized map (Figure 3). Its shape is elongated and could broadly be
classified as fern-leaf type catchment. As catchment is small so the time of concentration
of flood would be small i.e. flood waters will reach reservoir in a very short time.

Gross catchments area – Author has estimated catchments gross area by using computer
digitization method. Approximate original catchments area comes to approximately
1, 06,800 hectares. It is mainly Betwa and its tributaries excluding Kolansh river system.
When river Kolansh was diverted via upper lake, its catchments area up-to Bhadbhada
water-divide (39,658 hectares) got included, making the gross area as 1, 46,458 hectares.
The total area of five islands, situated in the Bhimkund reservoir is 1460.35 hectares.
Thus the entire contributing catchments is 147938.35 hectares

Kincaid (1888) reported area of Kolansh basin as 500 sq miles (129500 hectares) but this
figure does not match with current figure of GOMP or CWC (GOI). The current figure of
Kolansh basin is 36100 hectares and water spread area of upper lake is 3558 hectares.
The gross catchments area is therefore 39,658 hectares only.

Run-off contributions

Run-off contribution from all catchments - The gross run-off contribution has been
estimated by using rational formula assuming run-off co-efficient as 0.5 and annual
average rainfall as 1100 mm (1.1 meters). The estimated yearly gross run-off contribution
therefore could be 81366 hectare meters.

Ratio of catchments area and water spread area of reservoir - The ratio of water spread
area of reservoir to gross catchments area is 27 % 100 or 27% of the catchments.

Rain fall contribution of reservoir – As the reservoir area is very large so its contribution
has been estimated and added to the catchments contribution. The rainfall contribution,
over the reservoir (approx area 40020 hectares) is 44022 hectare meters. Seepage loss
from the lake-bed has been estimated by using Central Ground Water Board norms. It is
approximately 57.6 hectare meter per year therefore the net yearly contribution from the
lake is approximately 43964 hectare meters.

4
Gross contribution (reservoir and catchments) - Gross contribution is approximately
1, 25, 388 hectare meters only.
Storage capacity of Bhimkund

Storage capacity of the Bhimkund reservoir - It has been estimated by assessing

reservoirs area, side slopes, basement morphology and average depth. The reservoir
correction factor (0.7) has been applied to assess the gross effective storage capacity. On
the basis of above assessment, the effective storage capacity of Bhimkund reservoir
comes to approximately 5, 60, 280 hectare meters.
Relation between storage capacity and catchments yield - The approximate yearly
contribution is 1, 25, 388 hectare meters where as storage capacity is 5, 60,280 hectare
meters i.e. reservoirs capacity is nearly 4.5 times more. This relationship indicates that
initial filling of the Bhimkund reservoir might have taken nearly 5 years.

Shape and size of the reservoir

Author has attempted to assess the shape and size of the reservoir along with contributing
catchments by using computer digitization technique. The digitized map showing above
details is given in Figure 3 below.

Figure 3 Digitized diagrams of Bhimkund tank and its catchments

5
The study reveals that gross water spread area of the reservoir (area within blue
boundary) is 40022 hectares. The area of 5 islands, within the reservoir is 1460.35
hectares. Mandideep is located approximately in the centre of the lake. Other four islands
(not reported by Kincaid) are located in southern and eastern part. They have been shown
in dark yellow color. According to above map, the northern boundary of Bhimkund was
at Amravad Khurd (23º 12´ 30" and 77º 29´ 40" N) – a village located south of Bhopal.
The catchments area is within yellow but outside the blue boundary of the reservoir.
Note – The map of water spread area of Bhimkund is available in the report of Kincaid
(1888). Kincaid has reported the area as 250 sq miles (65000 hectares). Author’s findings
about water spread area and shape of reservoir are different.

Silt

Silt deposition, primarily depends on relation between inflow in the reservoir and out
flow from the reservoir. As Bhimkund survived till 1430 therefore, the silt deposition was
also confined till 1430 i.e. for nearly 400 years.
Catchments of Betwa river-system were the principle source of the silt for Bhimkund. Its
maximum quantity was brought to the reservoir during floods. The thickly forested
catchments with variable slopes and hills of moderate height, even during floods, would
have allowed movement of only small particle of weathered basalt and sandstone. The
silt, therefore, consists of soils derived from weathering of these rocks.
Study of bore-hole logs at 18 locations in the reservoir area indicates silt thickness. Its
maximum thickness is 28.06 meters (village Satalpur, east of Mandideep). Samples
indicate that the silt mostly consists of sticky yellow clay. The bore-hole data shows that
large particles (sand and gravel) appear along and near the buried river courses/shore line.
Bhimkund existed for nearly 400 years therefore it is concluded that silt deposition of
28.06 meter was confined to 400 years only. The maximum silt thickness of 28.06 meters
is abnormal. It is many times more than silt deposition in Upper Lake. This difference
probably hints the purpose of construction of Bhimkund.

Basin water diversion and river science

Traditional stories tell that to complete the number and water deficit, waters of a missing
river-system were diverted. Missing river has been named as Kaliasot. Kincaid (page
351) says that engineers of those days undoubtedly understood that the drainage area of
Betwa and its tributaries was insufficient for the purpose.

Influence of Bhimkund on Kaliasot river-system

Construction of Bhimkund has influenced the confluence of river Kaliasot with Betwa /
Bhimkund. The study reveals that river Kaliasot passed through three phases. These
phases existed in three different periods i.e. after the destruction of Bhimkund, during its
existence and prior to construction of Bhimkund. Author describes these phases as
current phase, intermediate phase and original or old phase. During these phases, river-
system had undergone major changes. The changes are as given below-

6
Current phase –The present day water course of river Kaliasot, from Bhadbhda (origin) to
village Bilkhiria Khurd (23º 8´ and 77º 27´ N), is unchanged. It is same for all the three
phases. The remaining part of the river course, between Bilkhiria Khurd to its current
confluence with Betwa has been developed in past 600 years on the lake-bed.
Intermediate phase –During this phase, Bhimkund reservoir was in existence. This phase
existed from the date of construction of Bhimkund to year 1430. In this phase, Kaliasot
traveled from its origin near Bhadbhada water-divide to village Bilkhiria Khurd. At
Bilkhiria Khurd, it merged in the Bhimkund reservoir.
Original or old phase – This phase is prior to construction of Bhimkund. In this phase,
Kaliasot originated near Bhadbhada water-divide and traveled via Bilkhiria Khurd to its
confluence (east of Bhojpur). In the initial reach, river flows through a structural valley
(broad at origin). The old river course, downstream of Bhimkund bund, can still be seen
by naked eye (Figure 4) on Survey of India topographic-sheet (55E/12). The original
drainage system of Kaliasot was developed after Deccan trap activity and is very old.

Figure 4, Waste-weir and old course of river Kaliasot

Influence of Bhimkund on Kolansh river-system

The traditional story talks about a missing river. The story tells that waters of this river
were diverted to another river basin. This river diversion has been studied in the light of
various probabilities including river piracy with the help of satellite imagery.
Satellite map (Figure 5,) of the area shows upper lake, lower lake, water divide between
Kolansh and Kaliasot drainage basins, many small streams flowing in respective basins
and river Kaliasot. In this map, few features of the imagery have been omitted.

7
Satellite map shows the origin of river Kaliasot and the extended arm of upper lake near
Prempura (within red circle). Kaliasot originates from a point (Bhadbhda area) which is
very close to water-divide of Kaliasot and Kolansh basins / watersheds.

The drainage basins of Kolansh and Kaliasot are a topographic reality. The water-divide
between them decides the slope direction. The slope in above drainage basins is away
(opposite) from each other therefore streams of one basin shall, as a rule, will always
flow in their respective basins and shall not cross the water –divide. The satellite map
(Figure 5) confirms above reality and rules out the possibility of river piracy i.e.
possibility of Kaliasot capturing the flow of river Kolansh. It denies the traditional story.

Figure 5, Satellite Map of Kaliasot River’s origin

The above mentioned reality over rules the possibility of river piracy (transfer of water of
Kolansh basin to Kaliasot basin). It also proves that traditional story of diverting the
missing river is scientifically unsound.

Upper Lake
Raja Bhoj constructed Upper Lake in Bhopal. It was constructed on river Kolansh which
was, earlier flowing beyond Kamla Park on the path carved by natural forces. Initially,
the water spread area of upper lake was 2.5 sq km. Traditional story says that upper lake
was constructed to divert the water of Kolansh basin to meet the deficit of Bhimkund.

8
Excess water of Upper Lake, which got released from Bhadbhda, followed the water
course of Kaliasot. Author is of the opinion that apart from Bhadbhda, the excess waters
were also flowing through Retghat boulder bund and the tunnel across Kala Park bund.
This assumption is based on the study of toposheet covering bund area.
The Figure 6 is a satellite image and shows upper lake, lower lake (blue area), river
Kolansh (light blue line), vegetation (red patches) and lineaments (light yellow straight
lines). Lineaments are weakness planes.

DRAINAGE & LINEAMENT OVER SATELLITE IMAGE OF UPPER LAKE BHOPAL

Geoscience & Geoinformatics Group , RSAC , MPCST, Bhopal

Figure 6, Drainage and lineament map of Upper Lake, Bhopal area.

The satellite image shows the river course of Kolansh on the Upper lake bed. It also
indicates that Kolansh River is flowing towards Kamla Park. The river path suddenly
disappears in the dark blue area (towards Kamla Park). The observations of river course
also indicate that there are no indications of river path diversion. It also appears that the
path is not influenced by lineaments crossing the upper lake.

Construction of Upper Lake blocked and submerged the route of river Kolansh. Thus, the
identity of river Kolans was permanently lost. The surplus waters of upper lake were
flowing through Bhadbhda, Retghat and tunnel across the Kamla Park bund. The release
of surplus waters has modified the flow pattern of Kolansh river-system and the same is
described in brief in the next few lines.

Impact of construction of Upper Lake on flow pattern of Kolansh river-system

Kincaid (1888) had written that the surplus waters of storage-lake (upper lake) were
carried into the larger lake (Bhimkund) for full three months after the close of rains.

9
Kincaid’s report is silent about the release of surplus waters from Kamla Park tunnel and
Retghat. The Kamla Park tunnel and Retghat were also the outlets from where the surplus
waters of upper lake got released. Since Kamla Park and Retghat were located on the old
river course of Kolansh therefore the released surplus water followed the same old course
and rejuvenated the old river. This new born stream of tenth century was however not
given any name.

Chote Khan (Minister in Bhopal Princely State) in 1794 (?) constructed Chota Talab
(small tank – original reservoir area 0.25 sq mile) on above stream but the surplus waters
of upper lake, after filling Chota Talab, got released and gave birth to a new stream. This
new stream is known as Patra. It flows on the old river course of Kolansh and meets
Betwa near Islamnagar. This is the brief history of influence of construction of upper lake
on the flow pattern of Kolansh river-system.

The waste-weir at Bhadbhada had also undergone changes in different periods to meet
water requirements. Nabab of Bhopal constructed stone masonry spill-way (approximate
height four feet). In 1963, GOMP constructed waste-weir, nearly 150 meter downstream
of stone masonry spill-way and increased the storage capacity. Similarly the height of
Retghat and Pulpukhta (Road Bridge) was also raised from time to time

Catchments character and Silt deposition in Upper Lake

Catchments of kolansh river-system were the principle source of the silt. It was brought
to the Upper Lake reservoir through surface run-off. In the tenth century, the forest cover
in the catchments of Kolansh must have been very dense. It must have been similar to
catchments of Bhimkund. The topography and the forest cover of the Kolansh catchments
would have allowed movement of small particle mainly derived from weathering of
basalt and sandstone.

The studies have been conducted to estimate silt thickness and its spread in the Upper
Lake. Studies reveal that near Kamla Park bund, two layers of silt are found. The lower
layer is stiff grey clay. Its thickness is between 2.50 to 3.00 meters and consists mainly of
clay and silt. It is hard. Above this layer there is another layer of soft grey clay which
consists mainly of sand, clay and silt. The lower layer is old and its period of deposition
is obviously till 1963. The upper layer is young. The silt deposition in Upper Lake till
1963 (before enhancing storage capacity by CWPC) is inferred to be between 2.5 to 3.00
meters. This thickness (2.50 to 3.00) is negligible or very small as compared to silt
deposition (28.06 in 400 years) in Bhimkund. The difference in the rate of deposition in
two reservoirs suggests that low rate of silting was for sustainability (long life) of the
reservoir while the high rate is for developing the undulating terrain in flat or gently
sloping ground.

Above study reveals that experts of traditional or indigenous water science knew-

1. Construction of nearly silt free water bodies. This was achieved by storing small
portion of catchments yield. This was further facilitated by keeping low or small

10
 ratio of reservoir area to catchment area. This is demonstrated in Upper Lake
where the ratio was kept as 1.61: 100 only. This ratio allowed maximum silt
deposition of 3.00 meters in 1000 years.
2. By retaining small portion of catchments yield and allowing removal of maximum
silt through waste-weir, they ensured pollution free water bodies. It may be
recalled that Upper lake started showing signs of pollution after 1945 when the
first filtration plant was established in Bhopal.
3. Construction of silt traps (water bodies) for reclaiming waste lands/ undulating
terrains. This was achieved by storing large volume of catchments yield. This can
be seen in Bhimkund where the storage capacity was designed five times more
than the catchments yield. This was facilitated by keeping higher ratio of reservoir
area to catchment area (27:100 or 27%). This design allowed maximum silt
deposition of 28.06 meters in just 400 years.

The above conclusions undoubtedly prove that people of Central in tenth century were
not only familiar with durable construction techniques but also knew the techniques of
accumulating silt in water bodies for reclaiming undulating terrains. They also knew the
construction of perennial and sustainable water bodies with inbuilt provision of silt
removal at no cost.

Selected References
Deshpande C.V., Study of water quality and quantity of upper lake using remote sensing
techniques., IIRS, Dehradoon, 1991.
Dey U.N., Medieval Malwa, 1965
Garg S.K., Hydrology and flood control engineering, 1993
Kand C.V., Indian experience in sustainability of civil engineering structures. ICR roving
seminar, January, 2010
Kincaid W. Rambles among ruins in Central India, 1888, The Indian Antiquary
Laurd C.E., Bhopal State Gazetteers. 1908
Paithankar M.G. and Saksena R.N., Some observations of the structure, tectonics and
geo-morphological features of the Vindhyan rocks of Bhopal, Madhya Pradesh.
Raghunath H.M. Hydrology. 1991
Rajia Hamid, Bhopal Darpan. 1998

*K.G.Vyas is a Former Advisor, Rajiv Gandhi Watershed Mission and Director,

WALMI, MP, E-mail: kgvyas_jbp@rediffmail.com

11