The site of the Triconch palace at Butrint lies on the southern side of the ancient city adjacent to the Butrint Channel within a slight re-entrant of the city walls. Previous investigations had revealed a late antique domus arranged around a peristyle. The 2000 excavations focused on the southern and western wings revealed 9 possible phases. Phase 1; in the early 4th century A.D. the waterside area was occupied by a number of structures including an arcaded building and bath house, some elaborately decorated as marble veneer was found in demolition debris. Phase2; in the later 4th to mid 5th centuries some of the structures were demolished and a new building erected over the top arranged around a stone flagged peristyle. The southern wing consisted of five rooms, a long gallery and a reception room, the latter two were paved with mosaics. The reception room contained a small marble-clad octagonal fountain. A small triclinium suite formed the western wing, three rooms which, through columnar screens, looked on to a fourth central room also paved with mosaic containing a Greek inscription. The wings fronted on to a portico which ran around all four sides of the peristyle courtyard. The column bases were white limestone and the columns red-pink marble. The portico in front of the triclinium was paved with a mosaic. Additions constructed in this phase included an apse to the reception room and a double apsed room to the south of the bath house. Phase 3; the mid to late 5th century new mortar floors were laid and a new larger tricilium (the dining room after which the Triconch was named) was built as an eastern wing, connected to the south wing by a double apsed vestibule and courtyard. The new northern range was just one room deep. Phase 4; in the later 5th the city wall was built around the Triconch. Phase 5; the rebuild was not finished and in the early to mid 6th century silt layers developed over the mosaics. The Triconch showed room divisions, door blockings, rough extensions and wall and roof demolition. Mortar built tanks, mussel shell layers, postholes, furnaces and fish hooks were found. The northern range was used as a mausolea. Phase 6; in the later 6th demolition horizons were interspersed with levels of domestic and commercial refuse, amphorae forming a major element of the ceramic assemblages. Phase 7; also in the later 6th century graves were created in all areas of the site. Phase 8; in the later 12th century a number of timber and earth buildings were constructed in the west triclinium associated with a well and layer of domestic refuse. Similar traces were found elsewhere. Phase 9; the topsoil was a black humic layer containing pottery from the 15th to 20th centuries.

The excavation during 2001 allowed the identification of three building plots that ran northwards from the Vivari Channel. The western plot is delineated to the south and west by the line of the city wall. The eastern boundary is thought to be adjacent to the small bath complex of the Triconch. In the 3rd or early 4th century a substantial building occupied part of Plot 1, indicated by the remains of an arcade built on a series of piers. Large quantities of transport amphora and access to the channel suggest commercial activity. Later a rectangular building was added to the arcade, which had most of its openings blocked. When the city wall was built earlier buildings were converted into a tower, which became part o of a gateway to the channel. In the early 6th century the arcade and early buildings were pulled down and shell middens were formed in the rectangular building. Other structures were built into the 7th century, which focused on the gateway. During the 11th to 13th centuries, Plot 1 was occupied by a series of timber buildings arranged around a metalled courtyard. Spreads of shell fish deposits and a series of hearths and ovens were found. Around the 13th century the gate was blocked and four crouched graves interned. The graves were disturbed and the large amount of soil indicates agricultural activity between the 14th and 16th centuries. Plot 2, contained the early domus, which expended in to Plot 3 when the large triclinium was added in the east. The 2001 excavations showed that this area was intened as a major monumental focus with entrances from the Vivari Channel and the city, and a substantial colonnade of marble columns and Corinthian capitals, fronted by a water-filled channel. Elaborate stone windows with Χ(Chi) Ρ(Ro) monograms were found.

This year’s excavations revealed more of the south facing corridor with its fine geometric mosaic that connected the reception room and the Triconch’s entrance way. The mosaic pavement was dotted with scaffolding holes and cement-mixer representing the rebuilding that was never finished. To the west, in Plot 1 further excavations showed that during the 4th century in the commercial area piers were inserted to support an upper floor. Destruction debris suggests the upper floor had polychromatic marble veneer. Flanking warehouses extended out from the building to the edge of the channel, which were incorporated in to the city wall when it was built. Towards the end of the 5th century the two storey house was falling into disrepair, as is attested by mid – late 5th century debris. In the abpve- mentioned remains was found an ivory gaming piece. In the 6th century the area was sub divided, an upper floor was inserted in to the western warehouse and the eastern warehouse was abandoned. Graves and features found during excavation indicted that occupation lasted until the 7th century.

This season saw the uncovering of the mosaic in the long southern gallery and its re-covering with a mesh and 30 cm of sand for protection until it could be conserved. The courtyard area was entirely excavated along with the northern portico of the peristyle. Beaten earth floors were found in the portico, where mosaics had been removed, supporting the idea that the rebuild was not finished. Several eastern rooms were also excavated including a small vestibule with a fine mosaic, which allowed access from outside the Triconch into the long gallery. The walls of the vestibule depicted a colonnade and a garden beyond. Excavations in the west of the Triconch showed that the colonnaded room of the earlier domus (phase 2) was blocked by a series of small walls. A massive threshold block to the north of the phase 3 triclinium indicated that the main entrance way had moved to there working in tandem with the double apsed vestibule. The vestibule, found in 2000, was built directly on top of the unpaved street, which ran along side the early domus. Excavation of the southern wing revealed that the westernmost room may have been used for domestic purposes. A large storage jar was found set in the floor and five complete oil lamps were found against one of the walls with two bone gaming pieces. Further burials including infant amphora burials were found. One of the rooms in the western wing was found to contain a circular structure which filled the room entirely and was probably used for storage. High denomination coins, including a gold coin of Basil II, from the 9th to 12th centuries were found, suggesting commercial activity. Further, postholes indicting several post built structures also suggested that the Triconch was not deserted during the early medieval period.