The zone modeled exhibited the behavior reported in Figure 1. Steam was injected into a well that had been hydraulically fractured for 310 feet. The injection pressure was gradually raised in an attempt to improve injectivity until the injection pressure exceeded 0.8 psi/foot of depth at the top of the perforations. The injection rate doubled immediately, and the temperature quickly rose 200°F over a hundred feet above the injection zone at the nearest logging observation well (LO1). The temperature logs reported in Figure 1 were taken six months later. The logs show that a high-temperature zone was growing up and to the west through LO1 and LO2. Figure 1 – Temperature Logs after 6 Months
This response was not seen to the north at LO3 nor to the east where there was no production well. Steam broke through in the upper zone of the producer 676 days after the fracture gradient was exceeded. It appears that the original injection fracture had extended upwards 150 feet, gradually rotated laterally and connected to one of the upper production fractures in the producer that was 135 feet west of the injector. Fracturing - The model of the north half of the pilot contained 2610 (26x9x10) grid blocks with variable dimensions as shown in Figure 2. After five years of primary production, the pressures in layers 3, 4, 6, 7 and 9 had been reduced substantially. The low pressure in layers 3 and 4 is especially important because this is above the steam injection zone in layers 5 to 9. As a result of both the low pressure above the injection fracture and the injection above the fracture extension gradient, the original fracture extended upward and westward. In earlier simulations fracture growth was modeled by manually changing the transmissibilities. Now, realistic parameters are used to predict that a fracture grew to the nearby well at the time reported in the pilot. The fracture growth was modeled using the PFRAC and DILATION options of STARS.
Figure 2 – Temperature
in a Growing Fracture in a Diatomite Reservoir
PFRAC allowed explicit prediction of vertical fracture extension (at the expected gradient of 0.6 psi/ft). When the gradient reached 0.8 psi/foot, westward growth of a fracture was predicted using the DILATION option. This specified “lateral” growth occurred in the same layers where the fracture had been observed to extend in the pilot. Diagenesis - As was just pointed out, diatomite is an immature silica rock that is subject to conversion into more stable rock when buried and heated. Normally, the process takes a few million years and results in a high permeability-zone containing many micro-fractures that eventually collapse to form a deeper, low-permeability, low-porosity shale. The process occurs rapidly when diatomite is heated. When the pore pressure is high, the micro-fractures that form when the matrix shrinks do not close and a permeable zone may be stable for some time. This process was modeled as thermal conversion of a component “Diatoms” into another component “Shale”, accompanied by both generation of carbon dioxide and permeability increases as high as a factor of 100. Figure 3 shows an example of combining diagenesis and fracturing.
Figure 3 – Diagenesis Increases Permeability
The figure shows that the permeability increased up to a factor of ten as diagenesis progressed. When used by itself, diagenesis could not explain the rapid movement of heat in the upper zone. Since there is a history of fractures growing laterally in shallow formations, it appears likely that horizontal fracture extension is the primary cause of the phenomena reported in this shallow diatomite reservoir. However, in deeper zones, which do not fracture as easily, diagenesis and silica dissolution could significantly increase permeability and affect the steam injection process.
This is the first demonstration of easily modeling diagenesis and
fracture growth. Earlier studies of this problem took months. This
study took a few days. Please contact MK Tech Solutions for more
information. |
||||||||||
|
|
||||||||||
|
||||||||||