Page 38

Page 39

Figure Above: Match of the entire DH4 data set with the LTT falloff model.

The above figure shows how the three numerical models considered above compare in terms of forecasts. The in-
jected volume for the wide model is seen to be essentially linear through most of the data, with end volume near 
73000 m3. The narrow model reaches a volume of 39000 m3 at the end.

Figure Above: Injection rate and volume forecasts from the mid-size model with a 
down-hole operating pressure of 130 bar.

8. Research

Summary of Injection and Falloff Data

 

    

- by Leif Larsen

DH4 Lower zone

The first test in DH4 in the Adventdalen Well Park was conducted with an active injection and falloff sequence Aug 
12-25, 2010, followed by an extended falloff with the last data registered July 25, 2013 – 25564 hours after shut-in. The 
last injection period lasted 121 hours at the rate 403 m3/d, which implies that the falloff was 211 times longer than the 
preceding injection period. The test was conducted in a 100 m open-hole low net-to-gross section from 870 – 970 m 
in the De Geerdalen formation of Late Triassic age. The tests involved water injection into a water-filled formation.

The following points summarize key observations and results from the DH4 lower zone data:

• Formation is under-pressured by at least 57 bar with the sea level as reference, with pressure 25 bar at gauge 

(853 m)

• Well clearly hydraulically fractured from the injection
• Not possible to determine permeability directly due to a lack of radial-flow data and also uncertain effective 

thickness

• Not possible to determine fracture dimension directly without known permeability
• “High” effective global permeability enhanced by a natural fracture network throughout the affected volume is 

possible with reduced lateral connectivity in the reservoir with sealing or partially sealing barriers parallel to the 
generated hydraulic fracture

• Properties in the high-permeability scenario: effective thickness 30 m, average permeability 2.26 md, fracture 

half-length 300 m, nearest leaky boundaries at 4.5, 15 and 26 m, model width 480 m, model length 6 km, radius of 
investigation along the fracture 3 km (3 years)

• “Low” basic permeability (matrix) is also possible with natural fractures only playing a role close to the generated 

fracture in a “stimulated” zone without large-scale flow barriers affecting flow

• Properties in the low-permeability scenario: effective thickness 30 m, average matrix permeability 0.12 md, frac-

ture half-length 450 m, permeability of “stimulated zone” 1.09 md, extent of “stimulated zone” 20 m (from fracture 
surface), model average width 1100 m, model length unbounded, radius of investigation 700 m into outer zone

• The main results from the analyses are qualitative with the suggested formation scenarios that can explain the 

pressure response from the lower zone of DH4 only treated as possibilities

Injection forecasts covering 1E5 hours (11.4 years) give essentially the same results for the two model scenarios, with 
rates dropping below 15 m3/d after about 3 years and total volumes 55000 m3 through the 11.4 years, but the areas 
affected by injection will be different. The results were generated with initial pressure of 25 bar and bottom-hole 
flowing pressure of 130 bar.