Fluid channelling to the surface
The case in Fig. 1 illustrates the detection of a downhole leak in an injection well. A leak in tubing, casing or a packer was suspected after detecting a high Annulus-B pressure of 820 psi and the appearance of oil at the surface during bleed-off.
HPT-SNL was performed in two modes: under shut-in conditions and during Annulus-B bleed-off that produced only 17 litres of oil and resulted in a pressure drop from 820 psi to 20 psi. As no pressure change was observed in Annulus A during the bleed-off of Annulus B, the 9 5/8-in. production casing was assumed to have no leaks. A static temperature profile (the red curve in the Temp column in Fig. 1) displayed no anomalies, and SNL data (the Shut-in SNL column in Fig. 1) also showed no noise.
During pressure bleed-off from Annulus B, low-frequency noise appeared in the interval between the dashed line and the surface (see the Bleed-off SNL data panel in Fig. 1). This noise pattern is characteristic of fluid flow through channels and fractures. Above the dashed line, both the static temperature profile (shown as the red curve in the Temp column in Fig. 1) and the bleed-off one (shown as the blue curve in Fig. 1) deviate, as the temperature sharply increased during Annulus-B pressure bleed-off due to the Joule-Thompson effect.
The survey prompted the conclusion that fluid from the upper part of an oil reservoir, overlying the injection zone, entered Annulus B through fractured cement behind the 9 5/8 production casing, which caused a pressure increase.
Fig. 1 Oil inflow in to B-annulus through channeling behind 9 5/8 casing.
This example demonstrates the detection of a packer leak in an oil-producing well. Excess pressure was found in Annulus A and a gas release was observed during pressure bleed-off of that annulus. The logging survey objective was to identify the cause of that annulus pressure.
That objective was achieved using the integrated HPT-SNL technique. The survey programme was designed to run temperature and spectral noise logging first under shut-in conditions and then during excess pressure bleed-off from Annulus A.
Shut-in SNL detected no fluid flow noise (see the Shut-in SNL data panel in Fig. 2). However, a temperature anomaly at the packer depth was recorded two days after shut-in (the red curve in the Temp column in Fig. 2).
The bleed-off temperature (shown as the blue curve in Fig. 2) is shifted towards lower temperatures, which indicates gas inflow from reservoirs below the survey interval. Additional cooling took place directly above the packer due to the Joule-Thompson effect. At the packer depth, SNL detected high-amplitude broadband noise, indicative of gas flow through the packer leak.
Fig. 2 Overpressure in A-annulus caused by packer leak.
Below is an example of tubing leak location in a gas-producing well. Increased pressure in Annulus A was found to be equal to the pressure in the tubing, which suggested communication between them. The objective was to locate the suspected leak. Gas inflow was detected during bleed-off.
HPT-SNL was conducted in two modes: under shut-in conditions and during pressure bleed-off from Annulus A. No noise was detected under shut-in conditions (as shown in the Shut-in SNL data panel in Fig. 3), which indicated that no flow was present in the well. For that reason, both static temperature (the red curve in the Temp column in Fig. 3) and SNL data were used as baseline profiles for leak location.
The temperature recorded during bleed-off from Annulus A (the blue curve in the Temp column in Fig. 3) was found to be higher than the static in the lower portion of the survey interval because of fluid flow from below. A pronounced cooling anomaly, generated by gas break out, was detected at the sliding side door (SSD), which was supposedly closed, according to the client. High-amplitude broadband noise was also observed at that depth, as shown in the Bleed-off SNL data panel.
The correlation of obtained survey results pointed at a leak in the SSD, which was claimed by the client to be closed. Gas flowing through the tubing entered the annulus through the SSD and created excess annular pressure.
Fig. 3 Gas inflow in to A-annulus through tubing leak.
The case below is an example of the detection of a production casing leak in an oil-producing well. The survey objective was to identify the cause of oil appearance at the surface and excess pressure in Annulus A. The objective was achieved by implementing the integrated HPT-SNL technique in two modes: under shut-in conditions and during pressure bleed-off from Annulus A.
Shut-in Spectral Noise Logging data revealed no fluid flow noise (see the Shut-in SNL column in Fig. 4). The static temperature curve also displayed no anomalies and was considered a baseline profile. SNL detected intense broadband noise during pressure bleed-off from Annulus A (see the Bleed-off SNL column).
Bleed-off temperature data (shown as the blue curve in the Temp column in Fig. 4) show a heating anomaly caused by the Joule-Thompson effect during fluid entry into Annulus A at the dashed line. Above that depth, the temperature curves recorded in the two modes deviate. Broadband noise detected at the dashed line is indicative of turbulent fluid flow through a hole. A combined analysis of data recorded at that level indicated a casing leak. Low-frequency noise above the dashed line was caused by fluid flow through Annulus A.
Fig. 4 Oil inflow in to A-annulus through casing leak.