THE Castle Peak power station blast which killed two workers last year could allegedly have been avoided if crucial testing procedures had been carried out with more care. This claim is contained in a Fire Services Department (FSD) report produced at the inquest into the deaths earlier this year. Attorney-General Jeremy Mathews has ordered a new inquest be held after the Sunday Morning Post revealed the jury may not have been given all the evidence before deciding the deaths of Yip Ka-pui, 40, and Wong Kwong-yu, 38, were accidental. In its conclusions, the FSD claimed China Light and Power (CLP) should have been more careful about the way it conducted tests on the purity of the gas in the hydrogen plant. Because of the testing method used, an inaccurate result was given. The report said: ''The sampling process and analysis did not reflect the true composition of the gas content in HPVs and other sampling points. ''If a true result had been available, this incident could well have been avoided.'' Under a section headed ''Observations of the Hydrogen Sample Taking and the Laboratory Tests by CLP'', it said: ''Discrepancy with the theoretical composition was found in CLP chemist's analytical results. ''It is suspected that the sampling and analytical procedures employed could have led to misrepresentation of the actual composition.'' Miscalculations were also allegedly made because incorrect pipe gauges were used. The report also casts doubt on CLP's explanation for the cause of the explosion, which also injured 19 workers. Conflicting details are contained within the power company's own report, a transcript of the inquest, and the FSD's extensive report. CLP's explanation for the cause of the blast was that air entered the system via two vent valves which were inadvertently left open. It has been suggested the hydrogen gas flow was blocked by the shut-off of another valve and the stoppage of hydrogen flow to the gas-holder. This version of events has now been discredited by the FSD report, ''An Investigation Report on the Cause(s) of Explosion''. Among the document's conclusions were: ''This scenario depends upon a complex chain of events and improper operations, and involves a number of assumptions which cannot be substantiated in evidence.'' CLP's version of events are thrown into further confusion by the accounts of two workers who appeared as key witnesses at the inquest. The Sunday Morning Post has obtained a full transcript of the inquest and in it both assistant charge engineer Ben Chau and plant charge engineer Paul Cheung stated the vents were not left open. Mr Cheung even goes so far as to say he personally closed the vents. The FSD report goes on to criticise CLP for a number of deficiencies. It says: ''This investigation had led to the revelation of a number of serious deficiencies of which those on August 28 were merely specific instances.'' A number of hours before the main blast, another ''bang'' was heard inside the hydrogen generation plant. The FSD report said: ''The team was disappointed that the 'bang' occurring at 0230 hours on 28.8.92 inside the hydrogen generation plant at the outlet of the water ring compressor was not recorded in the station log and also not investigated. ''The 'bang' was probably caused by an internal explosion in the pipework, if recorded should have alerted the management to a serious problem within the hydrogen generation plant.'' A number of alarms were tripped in the plant and, for some periods, were sounding almost constantly. ''The records provided by CLP were incomplete so it is not certain if any events with more serious consequence had happened,'' the report said. Earlier in the report, details were given of damage that was found on the gas holder. The damage could cause jamming of the inner cup of the gas holder. Two weeks ago CLP issued a statement claiming the jammed cup theory was ''discounted by all of CLP's experts''.
