Shortages of fresh water in southeastern Saskatchewan prompted the search for a frac fluid that would tolerate wide variations in produced water quality, without compromising stimulation results. (Images courtesy of BJ Services)

In many areas of the world, water shortages create problems for oilfield stimulation treatment plans. For example, dry weather, competition from agriculture, and rural populations limit the volume of fresh and surface water available for stimulation in southeast Saskatchewan, an area of significant Bakken Shale oilfield development.

The issue of water supply is significant not only because the development area is large but also because each well is typically a multistage horizontal completion requiring large volumes of treatment fluid.

One possible solution to the water supply problem is to use produced formation water or flowback water from prior stimulation operations. To enable this solution, a robust, flexible frac fluid system was developed — and has since been used to successfully frac more than 1,200 stages in more than 150 productive Bakken oil wells.

Bakken oil development throughout the Williston Basin is certainly not new. This resource has a 30-year history of vertical well development and the advancement of hydraulic fracturing technology. A major step change in exploitation of this vast oil reservoir is use of horizontal drilling and “toe-to-heel” multistage fracturing techniques using mechanical isolation technology.

The Williston Basin extends from South Dakota into North Dakota and Montana, and north into the Canadian provinces of Saskatchewan and Manitoba. Average depth in the development regions of Saskatchewan is 4,900 to 5,400 ft (1,500 to 1,650 m) and consists of three members: upper, middle, and the lower Bakken.

The primary completion target is the middle member, which is an argillaceous siltstone ranging in gross thickness from 0 ft in the northeast flank of the basin, to 80 ft (30 m) in the south. Water saturation averages between 40% and 60%, porosity ranges from 5% to 12%, and matrix permeability is 0.01 to 0.5 mD. This middle member is bound by the upper and lower Bakken shales. The upper shale is dark, organic-rich, fissile, and prone to natural vertical fractures. Its thickness ranges from less than 3 to 16 ft (1 to 5 m), but it is inadequate to completely control fracture height growth due its unreliable integrity.

Vertical well completion technology has proven to be quick and cost-effective. However, due to the wet, naturally fractured overlying Lodgepole formation, oil production was plagued with the quick onset of water production. Initial oil production for vertical wells averages 16 to 50 b/d (2.5 to 8 cu m/d) rapidly declining to nearly uneconomic rates while water production increases in the order of 62 to 190 b/d (10 to 30 cu m/d). Despite attempts at innovative completion techniques aimed at avoiding upward growth from any form of stimulation, this frustrating production profile has persisted and resulted in very poor oil recovery.

Horizontal drilling of wellbore trajectories exceeding 3,900 ft (1,200 m) combined with efficient multistage transverse fracturing are now routine and have quickly proven to be an excellent completion strategy. Further advancements include mechanical isolation of the openhole well bore and sequential fracture treatment of up to 13 transverse fracturing stages in the horizontal well bore. Historical fracturing fluids for the area have included crude oil, crosslinked refined frac oil, zirconium crosslinked water, crude oil-in-water emulsions, and linear to slickwater systems.

Due to the proximity of the water-bearing Lodgepole formation combined with the history of excessive height growth in vertical well fracturing, rigorous fracture simulation was performed. The simulation included detailed petrophysical analysis integrated with grid-based fracturing simulation. Ultimately, the treatments have been designed to place 8 to 10 metric tons of 20/40 proppant at low pump rates averaging 5 bbl/min (0.8 cu m/min). Fluid requirements average approximately 150 bbl (25 cu m) per stage, with total wellbore fluid reaching 1,900 bbl (300 cu m). Pump times for these treatments range from 40 minutes to 90 minutes per stage.

The low pump rates complicate the fluid chemistry for two reasons. First, the resulting fluid velocity complicates proppant transport, making fluid viscosity a critical issue, particularly with reservoir temperatures around 160°F (70°C). Second, the low rate and long pump time require the fluid to maintain a strict rheology profile until job placement is complete. Relatively cool reservoir temperature and even cooler fracturing fluid temperature challenge this predictability.

More recently, however, changes in water availability throughout the area put great pressure on operators and service companies to find new fracturing technologies that would not rely as heavily on freshwater.

The water conundrum

Regional geography consists of open prairie dominated by agriculture characterized by grain farming and domestic livestock. This area could be considered semi-arid, relying on annual rainfall to support all rural, agricultural, and oil and gas activities. Due to the rapid development of Bakken oil and subsequent water requirements, competition for suitable water has placed considerable pressure on the oil and gas industry to adopt new alternatives.

One obvious solution was to find a way to use produced water from this very active area. The task was not trivial. Water quality issues included:
• Salt content as high as 200,000 ppm. Salts are the biggest enemy to crosslinked frac fluid systems because they alter fluid pH, which prevents or delays crosslinking. Calcium and magnesium, in particular, are anathema to typical borate- and zirconium-based, crosslinked frac fluid systems. Post-frac, produced salts also can induce corrosion and scale deposition in tubulars and production field equipment.
• Iron content as high as 40 ppm. Iron affects crosslinked frac fluid viscosity by changing the oxidation state of the transition metal in the crosslinker or tying up crosslink sites to prevent proper crosslinking. This significantly affects the stability and rheology of the fluid. In addition, leftover Fe3+ can form sludge in oil, causing post-stimulation production issues.
• Residual corrosion inhibitors, iron control agents and other production-related chemicals. These contaminants can affect the stability of typical frac fluid crosslinks and also make it difficult to adjust chemical additives for a given operation.

In addition to water quality issues, a frac fluid system needs to be highly predictable during treatment operations, job placement, and shut-in period. Of the many variables that impact frac fluid stability, reservoir temperature is of primary concern. Knowing the extent and rate of fluid heat-up during the entire fracturing sequence is critical in designing the optimum breaker chemistry and loading.

In this area of the Bakken, typical reservoir temperature is near 160°F (70°C), but the observed flowback temperature at the wellhead is around 80°F (25°C). One might expect a large temperature change given the very long pump and subsequent shut-in times required to complete multiple stages in these horizontal well bores. Each stage can take up to 90 minutes to complete.

In fact, the reservoir in this area of the Bakken often acts as a temperature insulator: By recording wellhead temperatures throughout the entire flowback for one completed well bore, engineers found that frac fluids that had been pre-heated to 60 to 70°F (17 to 22°C) were heating to only 77 to 82°F (25 to 28°C) downhole.

Developing a tolerant fluid

Armed with a wide variety of produced water samples from the area, chemists in BJ Services’ Calgary analytical and stimulation laboratory developed a frac fluid system that would tolerate a host of produced water sources, despite large variation in chemical contamination and temperature fluctuations. To do so, the fluid system also would have to be crosslinked with stable rheology for sand transport and fluid leak-off control.

About a dozen produced water samples were obtained from producing formations in the area. Over the course of about two months, studies sought to understand the water chemistry of each sample and the performance of about six different polymers and combinations of crosslinkers and breakers with these water samples. BJ Services chemists were able to narrow down the target to a polymer family and, finally, to one high-performance polymer and crosslinker that were found to be tolerant of water chemistry. The final system choice generates in excess of 400 cp initial viscosity and remains stable and predictable for 100 minutes, at which time it undergoes a dramatic break.

During field operations, the system is easy to use. A quality control technician tests water quality, which determines any minor adjustments that may be needed in crosslinker or buffer loading. To date, the Viking II PW (produced water) system has been successfully pumped in about 1,200 stages in 150 Bakken well bores.

Production results have met every customer expectation for both gas and oil well applications. Further, the system offers operators benefits in terms of environmental stewardship and economics. One customer estimated that the new fluid saved 10% to 15% of total stimulation cost because of reductions related to hauling, heating, and disposing of fluids, as well as other costs.

The system also has been used in other fields, particularly in remote areas where suitable freshwater is costly due to the trucking cost. It has been used in several Devonian oil wells in west-central Alberta and in several Cretaceous tight gas wells in northwest Alberta. The system is currently being tested for use in North Dakota Bakken wells.

Comments