In the late 1880s, Lieutenant Governor Sir Walter Hely-Hutchinson instructed engineer Osbert Chadwick to conduct an inspection of the water supply in Gozo. At that time, Gozo was dependent on water from natural springs and water from surface flow; however, up to 1886, there was no evidence of great scarcity of water.

Chadwick remarked that, due to the geomorphology beneath them, the villages around Mġarr ix-Xini depended solely on surface water collected in tanks. He also reported that a number of springs in the vicinity of Għajnsielem were flowing into the sea near Mġarr harbour.

Sketch of the pumping station.

Based on his experience in Malta, particularly after the great success of the pumping station in Wied il-Kbir, Chadwick was confident that there was an abundant amount of good quality water beneath the deep valley of Mġarr ix-Xini.

Moreover, the abundant vegetation, carob trees, figs and vines led him to the belief that water was present relatively close to the surface of the valley bed. Any water obtained from this source could be used to supply the villages of Sannat, Xewkija and Munxar.

However, in the winter of 1887, rain was well below average, and as a consequence a water shortage followed in the dry summer months. The government was obliged to cart water to Sannat, Xewkija and Għarb to quench the thirst of the inhabitants.

As a result of these urgent circumstances an immediate countermeasure was required. Thus, in January 1888, resident engineer G.C. Schinas proposed a project by which the water from the springs in Għajn­sielem would be pumped to Victoria, Xewkija, Sannat, Munxar and Għarb.

This proposal was chosen over Chadwick’s as it had the merit of being based on data already ascertained without incurring the delay involved in trials required in the Mġarr ix-Xini project. Chadwick’s idea was dropped because until the nature of the rock had been tested by manually excavating deep shafts into ground there was no certainty of success.

Due to the geomorphological features of the site, the project entailed a major undertaking bearing in mind the limited means available in those times

However, almost a decade later, in 1897, the supply from the springs in Għajnsielem proved to be insufficient.

Hoping for a similar success that had crowned the operations in Wied il-Kbir, Chadwick again proposed that shafts be excavated in the valley of Mġarr ix-Xini on ground levels between 50 and 100 feet above the water table. Evidently, his proposals were accepted.

Due to the geomorphological features of the site, the project entailed a major undertaking bearing in mind the limited means available in those times. The trial shafts that were dug to assess the quality of water were then connected by a series of mines, or galleries, at water table level. These mines would, in turn, direct the flow of water into a pit from where it could be pumped into high level reservoirs, to be distributed by gravity for consumption.

Drawing illustrating a typical pump which would have been used in conjunction with the boilers.

Considering the relatively rudimentary technology available at that time, the pump had to be operated by a coal-powered steam engine and so, an interdependent infrastructure had to be erected. This consisted of a system of chambers, excavated in the south side of the valley.

These chambers were in turn connected to the pit receiving water from the mines and to the road to Mġarr ix-Xini through a number of stairwells and a shaft for lowering bulky items. Two further shafts where excavated – one was used to extract the engine exhaust, and was connected to a chimney, while the other was used in conjunction with a chute to lower coal into a storage bunker.

Internal chambers carved in coralline limestone.

A bridge was erected to act as a support for a pipeline which spanned across the valley and went up the sheer height of the creek through an excavated tunnel. This tunnel made its way to a reservoir in Nadur, a village built on one of the highest hills on the island. A secondary pipeline supplied water to another reservoir at Ta’ Ċenċ.

Between 1925 and 1927, a secondary 600,000-gallon engine was installed so as to ensure continuity of supply. It is believed that this engine was powered by a diesel generator; however this cannot be confirmed as yet due to lack of sufficient evidence.

To house this plant, an auxiliary chamber was excavated and linked to another chimney. This addition can be clearly identified as it was clad in clay bricks, a feature complete alien to the rest of the building. The underground galleries were also extended further by 170 feet so as to be able to extract larger amounts of water.

More than 20 years later, around 1949, the plant was replaced by two Harland electrically-driven pumps. It was also around the same time that a five-million-gallon reservoir was erected at Ta’ Ċenċ to cope with the increased demand for water after the war. On its own, this reservoir provided twice the storage that was available before its construction.

Around 1959 a new pumping station was erected in Xewkija and was connected to the same system of galleries, or mines, from which the Mġarr ix-Xini station extracted its water. It soon proved uneconomical to run the two stations when extraction could be centralised at Xewkija.

Internal rock carved chambers.

The problem was that part of the water in the galleries flowed towards Mġarr ix-Xini station while the rest towards Xewkija station. Thus, in 1960 it was proposed that the galleries be deepened by an average of two-and-a-half feet, so that all the underground water could flow towards the newly constructed station. This deepening of the galleries also increased the natural flow of underground water as a consequence of a larger surface area and a higher pressure head.

In a telegram it was indicated that the increase in electrical consumption that the Xewkija pumping station required to cope with the larger volume of pumped water was even less than half the labour costs saved by the closing down the Mġarr ix-Xini station. Thus the services of the last two engine drivers and two mates who operated the Mġarr ix-Xini station were no longer required and the structure was left abandoned up to this day.

The research project was carried out by architecture students Mario Pace, William Moran, Robert Pace, Andrea Zerafa, Julian Vassallo, Samuel Bonello and James Dingli, under the supervision of Paul Curmi from HSBC and Paul Micallef and Charles Camilleri from the Water Services Corporation. The students would like to thank R.A. & Sons Manufacturing and Sannat local council for sponsoring their work.