Integrated Approach to Well-Leak Diagnostics Meets Success Offshore Timor Sea

You have access to this full article to experience the outstanding content available to SPE members and JPT subscribers.

To ensure continued access to JPT's content, please Sign In, JOIN SPE, or Subscribe to JPT

The identification of low-rate leaks along with low annular-pressure buildup rates in any type of completion presents challenges to well integrity. This paper emphasizes the importance of understanding well-diagnostic problems to determine feasibility, isolate interest zones, enhance stimulation strategies, and optimize the acquisition of high-resolution acoustical data from the wellbore with a new leak-detection tool. This case study discusses the methodology that underlies the successful determination of the depths and the radial locations in the outer casing strings of multiple leaks in an offshore well.


The studied well in the Timor Sea offshore Australia was completed as a horizontal gas producer in 2010. The well featured a single packer completion with 7-in./9⅝‑in. tubing and a horizontal lower completion with a swell packer and preperforated 5-in. liner. The well contained a production casing (13⅜ in.) with an A/B annulus (7-in./9⅝-in. tubing×13⅜‑in. casing and 11¾-in./9⅝‑in. liners) and surface casing (20 in.) with a C annulus (13⅜×20 in.).

In late 2016, the pressure of the A/B annulus was observed to have decreased from 10 to 0 barg over a 4-month period, having previously been stable at 10 barg for 1 year. Multiple attempts were made without success to restore pressure by topping up the A/B annulus with inhibited water.

In October 2017, the pressure of the A/B annulus was observed to have a slight vacuum effect of –20 kPa. Diagnostic testing was performed to confirm the issue. The A/B annulus was topped off with 46 m3 of inhibited brine. While filling the last 4 m3, with the A/B annulus confirmed full, the pressure plateaued at 4.25 barg while it continued to pump at 145 L/min, which confirmed a potential leak in the A/B annulus. After the pump was shut down, the A/B annulus pressure stabilized at 3.32 barg, and the fluid levels remained full. The A/B annulus pressure, with the well on line, steadily increased to only 4.2 barg, further confirming the presence of a leak.

In December 2017, a well-integrity logging suite, consisting of an advanced array leak-detection tool and conventional production logging sensors, was conveyed on wireline to determine the A/B annulus leak location. The decision would be made to suspend the well on the basis of the logging results.

Data Acquisition. In December 2017, wireline data were obtained across multiple passes and stations. All logging sensors used in the well have the capability to record stationary measurements and record data while the toolstring is moving, either while uplogging or downlogging. When recording a depth-based pass with the leak-detection tool, a bandpass filter is used to filter the conveyance noise created as the tool moves; typically, this conveyance noise dominates the frequencies below 5 kHz. Conversely, when recording a stationary measurement, the full spectrum of frequencies can be analyzed.

During the logging operation, leak activation was achieved by using a diaphragm pump to flow inhibited water into the A/B annulus without exceeding 4.25 barg.

The acquired data can be summarized as follows:

  • A downlog from the surface to 4400 m measured depth (MD) with leak stimulation active (low pump rate)
  • An uplog across the packer zone with no leak stimulation
  • 15-minute stationary measurements at preplanned and real-time-identified areas of interest
  • 15-minute repeat stations at suspected anomalies

During each stationary measurement, levels of leak stimulation were varied to invoke a range of flow rates. Stimulating the well with different levels will induce flow and produce varying acoustic levels across the duration of the station measurement. Response by toolstring sensors to the varied stimulation is considered to be positive evidence for an anomaly. At each 15-minute station, the first 5 minutes were used to establish a baseline measurement without leak stimulation; then, the pumps were started at a low rate. After 2.5 minutes, the pump speed was increased for an additional 2.5 minutes to encourage increased leak stimulation. The remaining 5 minutes captured the log response without active leak stimulation.

Tool Theory

The advanced array leak-detection tool consists of eight sensitive hydrophones arranged in a 28-in. vertical array. These hydrophones detect pressure and confirm the vertical and radial position of a potential leak in the wellbore. It is important to distinguish that the tool is not measuring the pressure of the leak, but the pressures of the sound (compressional) waves in the borehole fluid caused by fluid or gas flowing through the orifice of the potential leak point. Although this technology can also determine flow rates, phase types, and flow mapping, flow rates and phase types are not part of the analysis presented in this paper.

Data are recorded with two different sample rates and listening times. The 5-ms data are compressed for electric-line telemetry and used to provide a real-time result in the field, whereas the 100-ms data (uncompressed lossless memory) are analyzed in the processing center, and can be used for further analysis with improved resolution.

Nyquist filters are applied to the data in the field to prevent any temporal aliasing. Power spectral density (PSD) spectrums are calculated from the acoustic amplitude data recorded by the hydrophones. PSD spectrums show how the signal power changes with frequency and are used to analyze complex signals. The PSD spectrum can provide the first indication of a potential leak by showing distinct signal power fluctuations with frequency resulting from the leak’s noise.

Both the radial and vertical position of an acoustic source can be estimated by using the spatial array of sensors with a beamforming process. Beamforming triangulation techniques, unique to the tool, use knowledge of tubular sizes and annular fill in combination with the array of hydrophone sensors. Phase shifts are calculated between sensors and used to triangulate the location of the noise source. The process is implemented in the frequency domain, and the result is a 2D color map in which the estimated acoustic source position is indicated by the highest energy value.

Interpretation and Results

While injecting water into the A/B annulus, an initial downlog from the surface to 4400 m MD enabled the determination of fluids within the wellbore. An oil/water holdup level at 2697 m MD and an oil/gas holdup level at 3128 m MD were identified in the tubing. This interpretation of annular contents can then be used in the subsequent leak-detection interpretation and flow mapping. In addition to fluid contacts, zones of interest were identified for temperature anomalies for further diagnosis using stationary measurements with the leak-detection tool.

Leak-detection-logging results were monitored and interpreted in real time by means of a remote Internet connection, which enabled reactive station selection and near-real-time interpretation. The next step in the interpretation methodology was to conduct stationary measurements across the temperature anomalies to determine whether acoustic anomalies were also present at these points, and the most-critical factor of determining radial locations of these events.

Acoustic Investigation of Temperature Anomalies. An anomaly at 875 m MD was investigated with a stationary acoustic measurement. The tool was stopped with the hydrophone array centered at 875 m MD. High-resolution data began recording at a sample frequency of 500 KHz for 15 minutes. During the first 5 minutes, the A/B annulus communication was not stimulated; after 5 minutes, the leak stimulation began by pumping fluids from the surface down the A/B annulus. After an additional 2.5 minutes, the pump rate increased to achieve the maximum leak stimulation. Increasing the pump rate at the surface had the desired effect of increasing the downhole fluid-loss rate, thereby increasing the acoustic energy at the leak point.

Within the advanced leak-detection hydrophone array, hydrophones are tuned to different gains such that a range of pressure-detection sensitivities exists across the array. Hydrophone 8 is the most sensitive in the array and has clearly detected the acoustic anomalies occurring at 875 m MD. When the leak stimulation is increased to its maximum rate, the tool detects a clear corresponding boost in acoustic amplitude. In this case, the temperature sensor also responds to the stimulation event. Although Hydrophone 8 identified the correlation between stimulation and acoustic amplitude with the most clarity, the other hydrophones detected the leak when the stimulation was greatest. Using the array of hydrophones and knowledge of the annular materials, it is possible to compute a 2D radial-flow map using the beamforming technique previously described.

Fig. 1 shows the 2D radial-flow map results, indicating a focus of energy across the outer casing string wall. The results indicate an interpreted leak point with greater clarity. The investigation then focused on the area of 1520 to 1555 m MD, which indicated a temperature anomaly during the downlog. The leak-detection tool’s hydrophone array was centered across 1555 m MD while the same stimulation exercised was performed. Acoustic data clearly responded to leak stimulation; the acoustic amplitude increased in response to a greater flow rate. The 2D flowmap computed from the hydrophone array data indicated a potential casing leak at 1555 m MD.

Fig. 1—2D radial-flow map calculated from hydrophone array at 875 m MD (x-axis=in. from the tool).


A possible hypothesis for the fluid losses within the A/B annulus is a potential failure of the packer at 4306 m MD. Consequently, this area was investigated with a moving pass across the packer area and a series of stationary measurements following the same principle as previously described. Although low-frequency background noise levels around the packer area appeared to be slightly higher than in the previously discussed shallower stations, no significant change occurred when the leak stimulation was applied; consequently, it was concluded that the packer was not leaking.

This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper IPTC 19448, “An Integrated Approach to Well-Leak Diagnostics: Case Study of the Successful Application of the Latest Leak-Detection Technology and Interpretation Offshore Timor Sea, South East Asia,” by Andrew Imrie, SPE, Brendon Negenman, Chung Yee Lee, Mahadevan S. Iyer, Sarvagya Parashar, SPE, and Mohamed Raouf Shata, SPE, Halliburton, and Sean Helton, SPE, ConocoPhillips, prepared for the 2019 International Petroleum Technology Conference, Beijing, 26–28 March. The paper has not been peer reviewed. Copyright 2019 International Petroleum Technology Conference. Reproduced by permission.

Integrated Approach to Well-Leak Diagnostics Meets Success Offshore Timor Sea

01 January 2020

Volume: 72 | Issue: 1

No editorial available



Don't miss out on the latest technology delivered to your email weekly.  Sign up for the JPT newsletter.  If you are not logged in, you will receive a confirmation email that you will need to click on to confirm you want to receive the newsletter.



No editorial available