The partners now believe they may have misunderstood the nature of the field.

Warro-5 and Warro-6 were drilled and fracced last year in a ‘quieter' area of the field, away from the deep-seated faulting that was blamed for the high water rates in Warro-3 and Warro-4, but with stubbornly low gas rates and high water flows the JV realised it had not grasped the complexities of the subsurface geology.

Warro-5 is averaging 500,000 cubic feet per day and 220 barrels of water per day, while Warro-6 is presently averaging 590,000cfpd with an accompanying water rate of 225bpd.

While the gas rates are lower than expected, both wells are flowing gas naturally, primarily from one zone, and early indications are that the water system is finite and water rates are likely to be manageable.

"We're learning as we go along," Keenihan said.

Perhaps one of the most important lessons so far is that the deep faults may not be as problematic as expected.

Warro-3 drilled across a fault and had high levels of gas and water production, while Warro-4 was so close to a fault it seemed to be the similar situation, but it now seems that may not be the cause.

There is actually a pervasive small scale fracture network throughout the whole gas accumulation, and that could be the source of the water.

"Right from the beginning of these two new wells we suspected there was a fracture network all through the field, because when you frac a well you change the characteristics of the rock," Keenihan said.

"We saw the results of the frac and we knew there must be some natural fractures because we were clearly not near any faults."

A natural fracture network can actually help the gas to produce to surface, and may be a boon, however the water levels need to be manageable using jet pumps.

Given the water rates are dropping, Keenihan believes the water may not be a long term issue, so the company has decided to re-test Warro-4.

He believes the Warro-4 test was too short and valuable information can be obtained about the potential of the upper reservoir section by retesting the well.

That well, which produced from about higher up in the thick gas column, flowed each zone for two weeks at a combined rate of 0.6MMcfpd with 650bwpd, but Transerv and Alcoa used an expensive nitrogen-based process to dewater the well.

Pending approvals they will retest the well with the same jet pumps that have successfully worked at the Warro-5 and Warro-6.

Warro-4 used larger scale fracs than the "short and stubby" fracs used last year order to avoid the likelihood of encountering any faults or fractures missed in the 3D seismic around Warro-5 and Warro-6, and the hope is that Warro-4 will provide indication of the long-term flow capacity of the upper sands.

With the accumulated data Transerv can establish a gas-water ratio, work out which zones are the best and then plan for the drilling and fraccing of a horizontal well in zone.

However, because horizontal wells are expensive Keenihan wants to be more confident about the subsurface geology.

"We are looking for more production history from these wells, test Warro-4, and then work out how to interpolate that to the other sands, and then think about what a horizontal well will get us," he said.

"The challenge is always going to be how do you add these zones up, and we don't know that. The second step is, if we can see the upper zones are the best, should we drill a more expensive horizontal well, and frac that.

"That will give us a 5-10 times increase in flow rates."

The existing wells could be commercial in the US, but in the Perth Basin, where there are few drilling rigs and higher costs, "the hurdles are just that much higher".

Overall, Keenihan says Transerv does not believe the sky has fallen in at Warro, where Warro-5 and Warro-6 have proven a gas column of 440m, for which the base has not been found.

"What we are seeing is that the positives are all still positive as far as the work we have done," he said.

"We have had a number of wins. We have shown we have a substantial gas column that goes on a long way, and we can produce gas from various sands from this column, but we do have to get on top of the water."

The company also believes that fracs no longer need to be constrained to structurally quiet areas in the field, opening up more drilling locations and the potential to use larger scale fracs.

The field, about 200km north of Perth, now has a contingent resource estimate of 2.5 trillion cubic feet of gas-in-place, with upside to 11.6Tcf GIP, making it Australia's largest undeveloped onshore gas field.

RISC says the total recoverable gas resource of between 2.6-5.4Tcf is far in excess of the 50-75 billion cubic feet minimum economic field size.

Keenihan said the remaining Alcoa farm-in carry was sufficient to cover Transerv's costs from the testing until August.