By Dr David Whitehouse BBC News Online science editor

Investigators have not yet officially concluded what caused the loss of space shuttle Columbia, but a consistent story is emerging even if many of the fine details are still unknown.

Gehman will look at Nasa

Most plausible is the scenario involving a breach between panels eight and nine on Columbia's left wing, allowing superheated gas to penetrate and damage the wing.

The damage site is consistent with the place where a suitcase-sized chunk of foam from the external fuel tank broke off and struck Columbia at 725 kilometres per hour (450 miles per hour) 81 seconds into launch.

Analysis is continuing, particularly testing of wing components, foam impact testing and scrutiny of Columbia's debris.

Meanwhile, the question of whether the US space agency (Nasa) needs changing has been raised, and if the agency really learned the lessons of safety, risk management and perception following the 1986 Challenger accident.

Not just 'what happened'

Harold Gehman, chairman of the Columbia Accident Investigation Board (CAIB), says that finding the cause of the accident is only part of the board's task. It will also look at the way Nasa is organised.

The left wing was breached

He says that the US Government has made it clear to him that they want more from his report than just "what happened".

It is already clear that the CAIB will produce a report that is more wide-ranging than the report into the Challenger accident in 1986.

Experts in risk analysis and how to address safety issues in complex engineering and managerial systems, say that the Challenger accident was not just a failure of the so-called O-rings on its solid-fuelled boosters. Likewise, they argue, the Columbia accident is not just a failure in its heat-protection system.

The problems of changing Nasa is clearly a complex task that Gehman appreciates is fraught with difficulty.

He has said the question must be studied thoroughly so that changes will not be made that have unfortunate and unforeseen consequences.

When is serious serious?

Central to the problem is what has been called the "normalisation of deviance".

Recently recovered shuttle debris

Experts have asked: with such a complex, and still experimental, vehicle as the space shuttle, how can engineers and managers asses, monitor and act upon problems and concerns?

And how do engineers and managers avoid inadvertently expanding the bounds of acceptable risk?

Some testimony to the CAIB has pointed out the problem with the philosophy of learning to live with damage to the space shuttle and then accepting it as normal.

In short, experts are asking just what is serious in a critical system.

On another front, Nasa is hopeful that the three remaining space shuttles can return to flight by early next year.

A resumption of flights would certainly ease the pressure of operating the International Space Station (ISS).

Without the shuttle the ISS is being resupplied entirely by Russia and the crew of the station is being reduced from three to two.

There are no immediate concerns but officials have said that resupply margins could be thin by the end of the year.