RFour Energy · Surveillance

GOWIS — Gas Oil Intelligence System

GOWIS is a field-level well surveillance application for gas and oil wells. It reads a production history, applies the standard diagnostic methods of production engineering to every well in the field, and ranks the field by what the diagnostics say needs attention first. An AI agent sits on top of the same computed results and answers questions about the field in plain language.

It is built for the routine that most surveillance work actually is: several hundred wells, monthly or daily rates, and a limited number of intervention slots. The methods below are the published ones — GOWIS runs them consistently across the whole field rather than one well at a time in a spreadsheet.

Diagnostic methods

MethodWhat it answersSource
Critical rate for liquid removal Whether a gas well is producing above the rate needed to carry liquids to surface, or is loading up Turner, Hubbard & Dukler (1969); Coleman et al. (1991)
Decline curve analysis Decline model, decline rate, EUR and remaining reserves per well Arps (1945)
Reciprocal rate method An EUR estimate that does not require choosing a b-exponent Reciprocal-rate straight line
Water control diagnostic Whether excess water is coning or channelling — different problems, different treatments Chan (1995)
Water-cut extrapolation Oil EUR from the water-cut trend when rate–time decline is unreliable Ershaghi & Omoregie (1978)

What the application does

Field overview

Every well is classified and ranked in one pass. Gas wells are scored against their critical rate; oil wells against water cut, water–oil ratio and gas–oil ratio trend. The ranking is the working list: which wells are losing rate, which are loading, which are watering out, and which are simply declining as expected.

Per-well diagnostics

Rate–time and rate–cumulative decline with automatic model selection, the log-log water-control plot with its derivative, water cut and ratio trends, and — for gas wells — the critical-rate curve plotted against actual rate so the crossover date is visible rather than inferred.

Reserves

Oil EUR is computed three ways — decline curve, reciprocal rate, and water-cut extrapolation — and the three are shown together. Agreement between them is the confidence statement; disagreement is the thing worth looking at, and usually means the decline is still transient or the water trend has not stabilised.

AI agent

The agent answers questions against the computed diagnostics — not against the raw data and not from a general model of how wells behave. Ask which wells are loading, why a particular well was flagged, or what the water diagnostic says, and the answer cites the numbers the tool produced.

On goodness of fit. A high R² on a decline fit confirms that a curve passes through the points. It does not confirm that the points describe boundary-dominated flow. A transient-contaminated history can return R² above 0.98 and an EUR that is wrong by a factor. GOWIS reports the fit quality and the flow-regime caveat together, because the first number without the second is misleading.

Access

GOWIS is an account-based application. Sign-in is required and new accounts are reviewed before access is granted. The full tool library lists what else is available, including material balance, nodal analysis, pressure-transient and rate-transient analysis.

Frequently asked questions

What is GOWIS?

A field-level surveillance dashboard for gas and oil wells. It applies critical-rate, decline-curve and water-control diagnostics across an entire well set, ranks the field by what needs attention, and provides an AI agent that answers questions about the computed results.

Which liquid-loading criterion does GOWIS use?

Both. The critical rate required to lift liquids out of a gas well is computed from the Turner (1969) and Coleman (1991) correlations, and the well's actual rate is compared against them. Turner's original correlation was derived with an upward adjustment to match field data; Coleman's later work, on lower-pressure wells, found the unadjusted form fitted better. Reporting both makes the disagreement visible instead of hiding it inside a single number.

How does GOWIS tell water coning from channelling?

By the slope of the water–oil ratio and its derivative on a log-log plot against time, following Chan (1995). A near-flat or gently rising derivative indicates coning; a steeply rising one indicates channelling through a high-permeability streak or behind casing. The distinction matters because the two conditions call for different remedial work.

Which decline models does GOWIS fit?

The Arps family — exponential, hyperbolic and harmonic — with the model selected from the data rather than assumed. The reciprocal-rate straight line and the Ershaghi water-cut extrapolation are run alongside as independent estimates of ultimate recovery.

What data does GOWIS need?

Well identifiers, dates, and production rates. Gas-well diagnostics additionally need wellhead pressure and tubing size to evaluate the critical rate. Everything else — classification, ranking, decline fits, reserves — is computed from that.

Related reading

RFour Energy · Tools · Notes & tutorials · About
Loading the application… if this panel remains, JavaScript is required to run GOWIS.