Ryder Scott model ensures gas availability for futures contracts

The model shows well sites and gas saturations, with the lowest saturations indicated in blue and green, which are areas of aquifer invasion. The complex reservoir is not amenable to a conventional evaluation due to the aquifer.

When the price of gas suddenly soars, the conference room at Tejas Energy LLC turns into a "war room." It’s the "reps vs. the techs" — two groups on the same team in a tug-of-war.

A marketer, eager to cash in on a gas-price spike, says he wants to open up wells all the way to quickly withdraw gas from underground storage so he can sell on the spot market.

Bill Peebles, director of reservoir engineering, says, "If we open ‘em up too hard, we may damage the gravel packs."

The marketer counters, "We’d make so much money that we can afford to break ‘em and fix ‘em."

Peebles abruptly says, "Put it in writing."

In a deregulated industry, gas-storage facilities have evolved from pipeline-operations tools to marketing tools. Against this backdrop, the marketing-vs.-engineering balancing act is played out every day in meeting rooms of the companies that own some 400 underground storage sites across 27 states in the U.S. The two sides recognize the validity of each other’s arguments.

"It’s true what our marketer said. We could have made enough profits on the contract purchase to more than cover the cost of fixing a gravel pack, which costs between $300,000 to $400,000," said Peebles. "Our management is well aware that we have to strike while the iron is hot and take our chances."

For Peebles, a one-person reservoir-engineering department assisting in operations management at the West Clear Lake gas storage facilities, his moderating influence is embodied in the Ryder Scott field model. "By referring to the model, we estimate the upper limits for withdrawal rates without causing well failures," he said, referring to built-in well controls that prevent excessive drawdown.

The two-phase, black-oil model is also an ally of the marketing representatives because it serves as a reservoir management tool by forecasting expected gas-delivery and -injection rates during cycling. The reps follow the production decline-curve and make sure they withdraw initial deliveries at a rate below the curve to ensure that they have enough capacity to meet their future end-of-the-month contractual obligations.

"The model gives us the confidence in production to allow us to get closer to the curve and sell as much as possible," said Steve Lichty, who handles business development at Tejas. "We have to meet delivery dates on (price) hedges and if we didn’t have the gas, we would suffer multi-million-dollar losses." He added that the model also enables the reps to understand what hedges they can commit to in the future.

Ryder Scott first performed numerical simulation on the depleted gas reservoir turned storage facility in 1993 to assist in evaluating the property before Tejas acquired it from Exxon Corp. West Clear Lake, near the Houston ship channel, is one of the largest gas-storage reservoirs in Texas.

"Exxon operated it to accommodate swings in the pipeline system. We now use the facility to store working gas to sell to our Houston ship channel customers and to expanded markets through the interstate pipelines," said Peebles.

This is not just the classic case of an independent buying a domestic property from a major and operating it more profitably and efficiently. Under deregulation, Tejas and other interstate pipeline companies are operating their storage facilities as business centers rather than transportation hubs.

As a result, great demands are placed on the technical staff to quickly anticipate maximum efficient rates for the wells in a commodity business that changes day to day. "There’s a large incentive to increase deliverability to meet instantaneous swings in demand. We may be injecting gas one day and withdrawing the next," said Peebles.

The complicated reservoir is not amenable to conventional evaluation methods, such as a P/Z material-balance analysis, because an active aquifer invades varying levels of the anticline Frio sand and acts as a lower boundary with the shale cap rock acting as a sealing boundary. So Ryder Scott engineer Tina Obut performed a history match of the reservoir-pressure behavior and water-level movement during depletion and cycling to incorporate in predictive simulation runs.

"Predicting when the wells would ‘water out’ was the most difficult challenge," she said. Ryder Scott engineer Tim Torres performed nodal analysis to evaluate the benefits of gravel-pack and frac-pack technologies on well deliverabilities. Tejas refers to the analysis estimates to maximize each well’s deliverability without causing excessive completion pressure drawdown and gravel-pack failure.

The frac-packs limited completion pressure drops and enabled well deliverabilities of more than 100 MMcf/D of gas. "The frac-packs have allowed us to produce at safe, yet high rates, allow gas to move more efficiently and keep the fractures open while controlling the sand," Peebles said.

Additional nodal analysis was used to evaluate bottlenecks in well-tubing strings, chokes and surface flow lines. Then the historical performance and nodal analysis were incorporated into the model.

Ryder Scott continues to update the model with additional data so that it becomes an even more accurate tool to evaluate the various strategies of maximizing deliverability and injectivity while minimizing pad or cushion gas volume through well recompletions, infill drilling and debottlenecking of wellbore and surface facilities.

Since initiation of the project more than five years ago, individual well deliverabilities have increased from 30 to 40 MMcf/D to 100 MMcf/D. Ryder Scott also modeled the addition of compression facilities that have increased available gas and operational flexibility.

"We can now withdraw 780 MMcf/D of gas, which is the maximum that our dehydration facilities can handle," said Peebles. "We analyze the addition of wells and surface facilities by referring to the model. Expansion is a continuous process."

Although there has been no recent period of sustained withdrawal to check the accuracy of the model, three years ago, Tejas did withdraw enough over a time period to compare the real deliverability curve and predicted one and Peebles remarked that they matched. "Following the model, we’ve had no well failures either, so it’s worked," he said.

Last year, Ryder Scott assigned another engineer, Miles Palke, to assist Obut in refining the model on a continuous basis. "I’m very happy with the work of Miles and Tina. When we’ve needed quick work for what-if scenarios, they’ve been very responsive and timely. Both of them can run the model, so if one is busy with another project, the other one takes over."

[About] [Calgary Office] [Client List] [Contact] [Downloads] [Financial Institutions]
[International Experience] [Newsletter] [Services] [What's New] [Home]

Copyright © 1998-1999, Ryder Scott Company.  All Rights Reserved.
1100 Louisiana, Suite 3800, Houston, Texas  77002-5218 USA
Office: 713-651-9191  Fax: 713-651-0849