ABSTRACT This paper describes a sub-sea hyperbaric Post Weld Heat Treatment (PWHT) system developed jointly by SHELL UK EXPLORATION & PRODUCTION and COMEX DIVING. It discusses:Assessment of power requirements and equipment,Initial tests conducted in the COMEX hydrosphere and in shallow water in Marseille,Full scale North Sea trials at -150 m, comparing the performance and results of two separate power sources and control units. Particular attention is drawn to the importance of the elements, insulation and controls used to achieve uniform distribution of heat at the required temperature in a HEUOX environment. 1. INTRODUCTION Following an extensive evaluation of the operating conditions of risers Shell Expro produced a specification for:HYPERBARIC GIRTH WELDING OF RISERS which laid down stringent sour gas service requirements, in mechanical testing, and in particular CTOD criteria. Results from hyperbaric weld procedure qualification tests in Marseille and atmospheric weld ability trials in Aberdeen demonstrated clearly that those requirements could not be consistently met "as welded". Procedures incorporating PWHT in both environments were qualified successfully. For comparison of "as welded"/PWHT, CTOD results, see Table 2. In order that we always retain the preferred option of welded tie-ins a joint group was set up to develop a hyperbaric PWHT system. 2. DEVELOPMENT PROGRAMME 2.1 Principle Used Following meetings held in London and Marseille, the decision was taken to design and develop a PWHT system based on experience of preheating and partial tests carried out from 1978 to 1980. The method selected was therefore that of indirect resistance heating using resistance heating elements wrapped circumferentially around the pipe immediately adjacent to the weld (one either side). 2.2 Objective of System The objective of the system was to bring the weld area on a 36" Øpipe in a hyperbaric habitat at -90 m to a stabilized temperature between 540°C and 580°C according to a defined heating curve. (See figure 1). Heating and cooling rates to be a maximum 175°C/hr. and soak time a minimum of 1 hr/in. of thickness. The heating pads selected were 200 mm wide (based purely on current density calculations and availability), set up one each side of the weld with a 25 mm gap giving a total directly heated band width of 425 mm (figure 2). A total insulated band width of 900 mm was prescribed for prevention of harmful thermal gradients beyond the heated zone. Maximum permissible variation of temperature during the heating and cooling periods was 150a°C. The requirement was that all temperatures be automatically and continuously recorded with sufficient thermocouples attached to demonstrate that all parts of the weld fell within the specified temperature range. 2.3 Pre-assessment of the Problem It was known from field preheat experience that for pipe in the horizontal axis position in air on surface needs more applied heat on the lower 3 to 6 to 9 o'clock sector than for the upper half of the pipe. This is due to the convection heat losses from the lower sector heating elements and pipe surfaces (exterior and interior) augmenting the conductive heat transfer of the upper sector heating elements.