Segment C

Segment C is located in the southern region of Norne Field, and contains four injector wells: C-1H, C-3H and C-4AH (water and gas), and C-2H (water). These injector wells were perforated before 1999, except well C-4AH, which started injecting in January 2004 to increase pressure for neighbouring producers and to enhance the oil sweep across the Ile Formation. Here, we observe variation in impedance mainly in the eastern region of segment. Figure 4.12a maps acoustic impedance differences below the top of the Garn Formation and Figure 4.12b maps these below the top of the Ile Formation. The areas surrounding the producer wells showed slight changes in impedance, while greater differences in impedance were observed around the injector wells (anomalies SC-1, SC-2 and SC-3).

Anomalies SC-1, SC-2 and SC-3 are mainly located in the eastern region of segment C (marked by red dotted ellipses in Figure 4.12), and we therefore focused on interpreting anomalies around the injector wells C-2H, C-4AH and C-1H. Figure 4.13a illustrates the acoustic impedance changes around injector C-2H. Anomaly SC-1 is an increase in impedance, indicating strong hardening of the signal across the Ile Formation. Since the hardening signal is concentrated around the well location and C-2H injected water from 1999 until 2006, anomaly SC-1 might be solely driven by water-saturation increases and not a pressure-related anomaly. Injected water (higher density) replaces

produced oil (lower density) and increases the P-impedance (Johnston, 2013) but, as seen above, a pore-pressure driven reduction in rock-frame modulus can dominate the impedance change. Furthermore, well B-2H started producing significantly more water after January 2004, highlighting that the detached area was partially flooded and the OWC rose to the upper part of the Ile Formation (see Figure 4.13a). Comparing the SC-1 anomaly with the simulated model, we note that the simulated pore pressure below the top of the Ile Formation remained almost constant (Figure 4.13b). Meanwhile, the water saturation in this region increased significantly from 2001 to 2006 (Figure 4.13c). However, the simulated water saturation spreads along the Ile Formation, and is not concentrated around injector C- 2H and producer B-2H, as in Figure 4.13a. We also note that injector C-2H completions were located in deeper layers, below the SC-1 seismic anomaly. We conclude that further investigation in this zone are necessary to better interpret the 4D seismic signal, especially at SC-1, and improve matching with the simulation model.

Figure 4.12. 4D acoustic impedance maps between the 2001 and 2006 surveys for (a) below the top of the Garn Formation and (b) below the top of the Ile Formation. Black lines represent the locations of the deviated producer wells, while dark blue lines represent the locations of the deviated injector wells; red dots denote well-head positions. In (a), the NE flank of segment C is probably water flooded around

the top of the Ile Formation.

Figure 4.13d illustrates the softening (anomaly SC-2) and slight hardening of the signal (anomaly SC-3) around gas and water injector C-4AH. Based on well information, C-4AH injected water into the Ile Formation until January 2005, which is a water-saturated layer in this region, and gas injection started in 2005, lasting 6 months. We therefore

conclude that the likely cause of anomaly SC-2 is a combination of increased pore pressure and reduced fluid bulk modulus, due to the 6 month injection of gas. It is probable that anomaly SC-2 is concentrated around injector C-4AH because gas injection was relatively short. The gas could have remained local to the injector, while pore pressure would have travelled much faster through the reservoir. This would explain the almost-uniform increase in simulated pore pressure for the entire region (Figure 4.13e) which is not localized around injector C-4AH. However, a local hardening effect, visible at well C-4AH below anomaly SC-2 (Figure 4.13d), led us to question the interpretation: as the local hardening lies in the aquifer, below the top of the Tilje Formation, a cause is difficult to identify and might therefore be due to a poor time-pick or a processing artefact.

Figure 4.13. Profiles of the inverted impedance difference and simulation model for segment C. Locations of the profiles are given by the broken lines in the thumbnail plot, and red dots denote well-

head positions. (a) A cross-section of the inverted impedance difference showing a strong hardening signal around the injector C-2H (black dots denote the completion of C-2H). (b) A cross-section of the

simulation model showing no pore-pressure change around anomaly SC-1. (c) A cross-section of the simulation model showing a significant increase in water saturation around anomaly SC-1. (d) A cross-

section of the inverted impedance difference showing anomalies SC-2 (softening signal) and SC-3 (hardening signal) around the injector C-4AH (black dots denote the completion of C-4AH). (e) A cross-section of the simulation model showing a gentle increase in pore pressure around anomaly SC-2. (f ) A cross-section of the simulation model showing the significant increase in water saturation around

We also cannot rule out that the hardening anomaly SC-3 could be a result of a significant increase in water saturation (Figure 4.13d) due to flooding from injector C-4AH reaching the Garn Formation. This interpretation agrees with that of Huang et al., (2013) but is slightly uncertain since the Not-1 Formation (the thin layer between the Garn and Aare formations) is known to act as a sealing layer due to its shale content. Possible explanations are that fractures in the Not-1 Formation may have developed due to overpressure during injection in well C-4AH, which carried injected water from the Ile Formation through the Not-1 Formation, or water encroachment from the aquifer in this region. Alternatively, the simulated water-saturation change (Figure 4.13f) suggests that water from injector F-4H in segment G has crossed the segment boundary, reaching the top of the Garn Formation around anomaly SC-3. The analysis shows that compartmentalization in the simulation model may still require reconsideration, even when there is a good match with observed anomalies.