Hong Kong Water Supply – Tai Tam Upper Dam (formally Tytam Reservoir)

Tymon Mellor: Following the completion of Pok Fu Lam Reservoir in 1871, the City of Victoria (modern Central) had a secure water supply throughout the year. However, within a year it was already recognised that the supply would not meet the growing demands of the population. A number of proposals were prepared to develop the Tai Tam catchment and supply fresh water to Victoria’s citizens, but the project was considered unnecessary until the British military complained about the state of public hygiene. The Government was then forced to take action and to start providing sufficient clean water to address the public health issues.

New Water Source

At the time of the commissioning of Pok Fu Lam Reservoir, the estimated daily water need per person in the UK was 19 gallons, which included for domestic uses, sanitary purposes, watering, and waste[i]. The total supply available in Hong Kong provided a daily equivalent of 5.75 gallons per person, and which reduced with a fixed supply and a growing population. In 1873, Mr J M Price, the recently appointed Surveyor General of the Public Works Department, was tasked with finding a long-term solution for the city’s water supply.

The problem was simple. Hong Kong had sufficient rain, exceeding even the greatest supply demands of the time, but it needed to be captured in a reservoir of sufficient size so as to hold enough water for a dry season, and at an elevation to allow the water to flow under gravity to the residential districts below. Two other issues were also important; locating the reservoir on the same side of the hills as Victoria so as to avoid the need for a tunnel or steam pumping engine; and finding a narrow gorge to minimise the length of the dam in order to minimise the construction costs.

A detailed topographical survey was undertaken over several months of the Hong Kong hills, confirming there were no ideal locations for the reservoir and only three valleys were of sufficient size to impound the necessary water. These were Pok Fu Lam, Aberdeen and Tai Tam, or Tytam as it was recorded at the time[ii]. Detailed schemes were developed for each location to establish the preferred arrangement.


Upgrade Pok Fu Lam

At Pok Fu Lam, two new reservoirs were proposed, one above and one below the existing structure. The combined capacity of the three dams would be 184 million gallons, but the estimated water supply for the valley from the surrounding hills was only 180 million gallons. The cost of the project was estimated at £201,500 and given the minimum requirements of the city was estimated at 330 million gallons, the site was dismissed.


Aberdeen Valley

Above the village of Aberdeen, a large valley, flanked by precipitous hills narrowing to a gorge 150m wide was an ideal location to build a dam. The dam structure would be 27m (90ft) high and 160m (520 ft) long, impounding 213 million gallons. The gathering grounds could yield 270 million gallons a year, ensuring the dam could be filled. The down side of the location was the reservoir’s elevation. With a discharge level of 73m (240 ft), it could only supply water to residents living below this level. Taking account for losses, this would limit the supply to Hollywood Road and below. The cost of the scheme was estimated at £250,000 but to make it a practical proposal, additional cost would be required for service reservoirs and pumping stations to lift the water to a higher level.

Given the limitations of the site, a scheme was developed utilising a steam pumping engine to transfer water to the existing Pok Fu Lam reservoir, and to utilise the existing distribution network. However, this approach would not address the main problem of supply to the eastern portion of the city. The scheme was not developed but the site was later developed for a reservoir in 1893 by the Tai Shing Paper Mill[iii].


Tai Tam

The study then identified the Tai Tam Valley, but there was a problem. The valley was on the wrong site of the island and would require a tunnel or long contouring conduit with short tunnels at the hill-spurs overlooking Deep Water Bay and Repulse Bay to convey the water to Victoria. The Tai Tam Valley with a 122m dam (400ft) would hold 300 million gallons, within a catchment that could supply 316 million gallons of surface drainage water. With 150,000 gallons a day from tributary streams to fill the reservoir, the new supply would be sufficient to increase the water supply from 5.75 gallons a day per person to the desired 19 gallons a day per person for the 125,000 population of Victoria. The dam would be located at an elevation of 122m (400ft), allowing the gravity supply of water to nine tenths of Victoria.

The final 1873 proposal was for an earth dam with puddle clay core to hold 250 million gallons, with a conduit winding around the hillsides, intercepting a further 25 mountain streams while transporting the water to Victoria. This would increasing the water supply to 13.5 gallons per head in the winter and 30 gallons per head in the summer.[iv] A service reservoir would be constructed in town along with a complete remodelling of the street distribution system for an estimate of £258,000.

Scheme Reduction

In 1875 the Colony was experiencing a severe economic depression, requiring the Government to allocate the remaining limited resources to the re-construction of the Praya Wall that had been destroyed the previous year in a typhoon. A number of cost saving measures were explored for the proposed Tai Tam reservoir including, reducing the size of the dam, changing the conduit to an iron pipe, cancelling the service reservoir, and delivering the water to the existing distribution mains in Bowington (Causeway Bay). Two schemes were developed, a ‘high level project’ avoided the need for a dam by intersecting the water source at around the 152m (500ft) level and transporting the water through a tunnel and conduit to the Wong Nei Chung valley, before using a syphon to drop down to the Happy Valley race course, and along Queens Road East in cast iron pipes, before connecting into supply mains in the high districts of Victoria. The Tai Tam stream could yield a possible 300,000 gallons a day during the dry season, or a little over 3 gallons per head. The scheme provided no storage, and would not supplement the water supply in the event of drought. The scheme was estimated at £39,085, a significant saving but failing to provide the volume and security of water required.

The second scheme was a ‘low level project’. This arrangement was based on the 1873 scheme but adopted a tunnel through the hillside and replaced the conduit around the hillside into Victoria with a cast iron pipe discharging into the existing Tank No. 1. The cost of this scheme was estimated as £122,600.

The change from conduit to cast iron pipe was not simply a change of construction form, it also had significant operational implications. The pipe would rust and would need to be replaced, but more importantly, the conduit would allow other passing streams to be fed into the system and the conduit provided storage of up to days demand. The conduit would be constructed as a series of flat narrow tanks, each around 400m long with a 300mm drop in between them rather than to a fall. Each water compartment could be converted in an emergency in to temporary storage through the use of stop-planks at each end.

Given the shortage of capital, in 1876 the ‘high level project’ was approved by the Secretary of State, Mr Gardiner Austin, and the following year, £50,000 was allocated to bring the Tai Tam water into town in a cast iron main without the provision of a dam.

Lethargy and Public Health

Despite the project endorsement, the scheme did not progress as there was little support to spend money on water supply. As reported in November 1878 Gazette[v] the Governor Sir John Pope Hennessy noted that with an estimate of over £300,000, “Her Majesty’s Government did not wish to have that amount of money spent in the Colony on additional water supply and for very good reasons. The Honourable gentleman referred to the drainage of the town. Even in one of our Ordinances, No. 8 .of 1856, I find that the Surveyor General is entitled to insist that every house in this Colony must either have a privy or water closet, and we are told in this report the advantage of the water supply would be immense owing to its utility in sweeping away the house sewage. Now the water supply that might be so used, that might be used if the Ordinances of the Colony requiring water closets were put in force, is it a water supply the Chinese have asked for? No. What do the Chinese themselves think of the proposal? They have told me what they think about it…They say, ‘Nothing alarms us more than your project of drainage and water supply for flushing house sewage. They are not consistent with our modes of living.’”

By 1882 the problem of the poor sanitation within the community and the poor public health was exposed to the British Government. There was concern that the poor sanitary conditions were impacting the health of the Colony’s troops[vi], “as Lord Kimberley says, by the office in command of the troops as to the condition of the Chinese houses, and the system of drainage, which he imagined was injuriously affecting the health of the troops under his command”. A detailed report was undertaken by Mr Osbert Chadwick, the Sanitary Commissioner and son of Sir Edwin Chadwick the motivator behind the UK Public Health Act of 1848. The report on Hong Kong was published in December 1882[vii] and resulted in the resurrection of the Tai Tam reservoir.

Chadwicks Report

Scheme Redesign

With an improved Government financial position, and the attention of the Colonial Office, the full scheme was resurrected for implementation. In discussions with Sir Robert Rawlinson, the Consulting Engineer for the Imperial Government, it became clear that the proposed earth dam was no longer suitable and needed to be replaced by a granite masonry or concrete structure. The reason for this change was the development in the understanding of the behaviour of large earth dams following a dam failure in Sheffield, England. On the stormy evening of 11th March, 1864[viii], the recently constructed Dale Dyke Dam located to the north west of Sheffield, failed. It released an estimated 3 million cubic metres of water, destroying 600 homes and killing 240 people. It was concluded that a leak had occurred around an embedded pipe, allowing water to penetrate the clay core washing away the heart of the dam and resulting in its collapse.

Great Sheffield Flood

The consensus of the engineering professions at the time and of Professor Macquorn Rankine [https://en.wikipedia.org/wiki/William_John_Macquorn_Rankine], responsible for the development of lateral earth pressure theory, concluded that where water would exceed 30.5m (100ft) the stability of the dam could not be assured. With a 27m (90ft) dam, with the option to extend at some future date above the critical 30.5m level, the 1873 proposal was compromised. Discussion in London between Mr Chadwick and Mr Rawlinson concluded that an earth embankment was unsuitable and a concrete structure should be adopted.

The new Tai Tam concrete dam design was for a 33.5m (110ft) structure, impounding 300 million gallons of water, with an estimated yield of 130,000 gallons a day[ix]. The dam was also designed to accommodate a further vertical extension of 3m (10ft) (added in 1895, raising the capacity to 390 million gallons)[x]. The conduit water supply arrangement was resurrected along with a storage tank and filter beds on the hill side of the Albany Valley (the top of Garden Road). Despite all the changes, the estimate for the works had been reduced from £259,000 to £166,440.

Tai Tam Scheme

Under the supervision of the Resident Engineer Mr G James Orange, construction of the Tai Tam water works commenced in 1883 with the issue of tenders for the construction of the tunnel and the foundations for the dam, along with the procurement of construction equipment including wooden huts to house the compressors required to support the tunnelling operation.


The Tunnel

Work on the tunnel commenced in April, 1883 at the Tai Tam end, and July, 1883 at the Victoria end. The 2,225m (7,300ft) long tunnel was to be constructed from both ends using a drill and blast method[xi]. Compressed air drilling machines were used to drill holes into the granite, which were then filled with dynamite to blast the rock. The tunnel, 2.1m (7ft) wide by 2.3m (7ft 6in) high, provided sufficient space to allow two sets of drills to operate concurrently. During construction, there was a high turnover of labour due to “Hongkong fever”, typically with 25% of workers incapacitated during the summer months, and during one period out of the 600 ‘coolies’ working on the tunnel, more than 420 were unable to work due to fever. Difficulties were also encountered with poor ground conditions and large inflows of ground water, but the biggest challenge was the rock.

Tunnel Section

Tai Tam Tunnel Section

The drilling was undertaken using a 2.5” Eclipse drill, powered with compressed air from a Hathorn & Co horizontal compressor, driven by a 12 HP steam boiler at each end of the tunnel. This set-up proved insufficient for the hard rock with slower than expected progress. In February 1885, an additional £4,000 was invested on upgraded equipment including, a new 30HP boiler, bigger compressor and eight new 3” drilling legs to replace the smaller units. With this change, production increased from 18m (60ft) per month to 24m (81ft) per month.

Hathon And Co

The two drives met on the 15th August,1887 after 47 months of tunnelling, with an accuracy of within 21mm (7/8”) and an expenditure of 28 tons (62,395lb) of explosives.



The Dam

With the abandonment of the earth dam scheme, a masonry solution with an outside curved face was adopted, but with the limited labour skills available, it was not possible to construct the works without excessive supervision, and thus it was decided to adopt concrete for the construction of the dam.

Tai Tam Dam, Construction Of The Base, Source HK Library From Tymon Mellor

Construction of the base of the dam 1884. Source: Hong Kong Library

The site of the dam was located where the granite bed rock was exposed in the stream. However, as the site was cleared, a number of joints, fissures, decomposed granite and a dyke were exposed within the rock. These features had to be dug out to a depth of between 3m and 4.5m and backfilled with a cement rich concrete to ensure the stability of the dam foundation. Drainage channels were constructed to lead spring water to the outside of the dam and to avoid a build up of water pressure below the dam. Tests were later undertaken and successfully confirmed the integrity of the foundation with water impounded behind the dam. Further testing indicated that the maximum daily water loss from the reservoir was not significant and most seepage was found to be from water seeping through joints in the rock, to the sides and below the dam, remote from the foundation.


The base of the dam was 19m (62’6”) deep and the top 7m (23’ 6”), constructed from a rubble-concrete core with a masonry inner face of cut stone and the outer of rough masonry to a height of 29m (95ft) in nine lifts. The rubble-concrete was formed by inserting rocks into the soft concrete then surrounding them with fresh concrete, to a ratio of 60% concrete and 40% stone. The concrete was hand mixed and rammed into place with iron rammers and from the feet of the passing “coolies”. Towards the top of the dam, as the working space reduced, a concrete mixer was purchased from the Kowloon Dock Company powered by portable steam engines. Steam cranes were used to lift the facing blocks into place.

A possible extension to the dam was identified with the construction of a small dam adjacent to the by-wash, forming a 20 million gallon reservoir. The cost of the new structure was estimated at £5,000, but this would not get developed until 1904.


The Conduit

A masonry conduit was constructed along the mountains from Wong Nei Chung to a new service reservoir and filter-beds in the Albany Valley. The path along the top of the conduit would later become Bowen Road. The conduit generally consisted of a channel 900mm (3ft) wide by 0.76m (2’ 6”) high with a concrete base and brick sides, covered with a 150mm thick granite slab to form a path. The capacity of the culvert was 7 million gallons of water a day. The conduit was built in level lengths, 230m long with a 150mm drop between each section. The conduit was 4.8km (3 miles) long and included 8 aqueducts of more than one span, the longest being 21 arches, along with 98 bridges and culverts.

1910s Bowen Rd Conduit

As the conduit passed by Wanchai Gap, a small service reservoir was constructed in the stream ravine. With a water depth of 7.6m (25ft) the reservoir had a capacity of 1.5 million gallons for a total cost of £3,650 including the cost of the diversion of the mountain road.

Wanchai Service Reservoir

Lime mortar was used throughout the conduit. The lime was made by burning and crushing sea-shells, then mixed, one-part lime to three parts of the local “red earth”, to form a sticky material that sets hard with time. To give the material added strength, a little cement was added, particularly when working in the winter. For bridges spanning up to 3m (10ft), 300mm square granite beams were adopted, beyond that stone arches were constructed. These utilised a rubble lime concrete, faced with stone. The arches were constructed from granite stones, typically 300mm square, set on a timber form until the arch was complete and stable. Cement grout was then used to fill all the joints.

Work on the conduit commenced in January, 1885 and was completed by May 1887 at a total cost of £34,700.

Albany Service Reservoir and Filter Beds

The new service reservoir and filter beds were located at the head of Albany Valley. The open service reservoir was set in the bed of a steep mountain stream, requiring the site to be enlarged to accommodate the 46m (150ft) wide by 9m (30ft) deep reservoir, sufficient to hold 6 million gallons. The existing stream and flood water passed below the structure in a 4.8m (16ft) culvert.


The dam wall for the reservoir was 4.5m (15ft) thick at the base, rising at 1 in 3, constructed of rubble-concrete faced with rough stone. The material excavated from the reservoir was tipped against the dam wall to provide additional strength and the outlet culvert was used to form the Garden Road bridge across the ravine.

Albany Filter Beds

During a rain storm on the 29th May 1889, a land slide trigged by construction debris from 36m above the reservoir, released 20,000m3 of material down the ravine, along with 200m of the Peak tramway. The reservoir was filled with the debris, including boulders over 200T, but none of the masonry or concrete structure failed. The debris was removed and the facility resumed operation.

1889 Albany Road Landslide

Like the service reservoir, the filter beds had to be cut into the hillside, requiring a cutting up to 18m (60ft) deep. Six sand beds, of an average area of 128m2, suitable for processing 650,000 gallons a day were constructed. These sand beds removed particles from the water and the associated microbes made the water safe to drink. Each sand bed had a filter layer at the base, covered with 800mm (2’ 9”) of washed sea sand. A 460mm (18”) water pipe connected the facility to the supply-main along Garden Road and into the city.


The cost of the filter beds and services reservoir was £22,761.


On the 1st November 1888 the new Tai Tam water works commenced supplying water to the city of Victoria. The final cost of the project was £209,579. Within six years, the water supply network was found to be insufficient and additional capacity was required. It was time to start constructing catchwaters and more reservoirs.


[i] Surveyor General’s Report On The Tytam Water-Works, 1st November 1885

[ii] Surveyor General’s Report On The Tytam Water-Works, 1st November 1885

[iii] Report on the Progress of Public Works During the Half-Year Ending 30th June, 1897

[iv] John M Price, Public Works Department Survey General’s Office Report, 10th July, 1875

[v] Government Gazette 23 November, 1878 page 565

[vi] Address of his Excellency Sir John Pope Hennessy, KCMG to the Legislative Council of Hongkong , 7th February 1882

[vii] Mr Chadwick’s Report in the Sanitary Condition of Hong Kong; with Appendices and Plans, November 1882

[viii] Great Sheffield Flood, https://en.wikipedia.org/wiki/Great_Sheffield_Flood

[ix] Surveyor General’s Report on the Tytam Works, 25th June, 1884

[x] Report on the Water Supply of the City of Victoria and Hill District Hongkong, Francis A Cooper, Director of Public Works 9th May, 1896

[xi] Tytam Water-Works, Hong Kong, James Orange, ICE Paper 2429, 1890






This article was first posted on 27th February 2019.

Related Indhhk articles:

  1. Hong Kong Water Supply – Mint Dam and Other Early Structures
  2. Hong Kong Water Supply – Pok Fu Lam Reservoir
  3. Hong Kong Water Supply – Kowloon Peninsula
  4. John MacNeile Price, Surveyor General of Hong Kong, the Tai Tam reservoirs
  5. The Tai Tam Reservoirs – recent photographs
  6. Tai Tam Upper Reservoir – Historic Building Appraisal
  7. Tai Tam Water Works – 1885 HK Government Report

Our Index includes many articles about reservoirs in Hong Kong.


Leave a Reply

Your email address will not be published. Required fields are marked *