Figure 2. Proprietary cement stress modeling software analyzes a 15.8-ppg FlexSet cement system (left) and a 16.4-ppg system. Although the heavier system had higher compressive and tensile strength, the software predicted it would not perform as well as the lighter system under wellbore conditions.

Salt has long created challenges for cement systems, but variations in temperature gradients, formation pressures and very thick salt sections in deepwater wells increase the complexity of providing long-term isolation and casing protection. The most significant problem salt poses for the industry is its tendency to creep. For example, salt can creep in on the hole during drilling, encapsulating and eventually (breaking) the drill string or narrowing the hole above the drill so that it becomes stuck. Creep may occur on a casing string over many weeks or even years, in the worst case deforming it so that it collapses the production string, requiring an expensive workover or well abandonment.

Salt creep

The rate of salt creep depends on depth, temperature, mineralogical composition, water content, formation impurities and differential stresses on the salt body. For example, pure halite (NaCl) is very slow-moving; high-water-content carnalite (KMgCL3•6H2O) and tachyhydrite (2MgCl2•12H2O) are highly mobile. Mobility increases with temperature; wells through formations with slow or no salt movement can experience sudden, catastrophic failures if producing the well results in increasing the formation temperature (Pattillo et al., SPE 109882). Therefore, any potential salt loading must be defined and considered in designing well casing, drilling and wellbore construction operations. The two-pronged goal of the cement job in salt formations is to isolate the formations and to support the casing so that the salt loading is uniform. This allows casing selection based on the overburden pressure, rather than some extreme potential salt load (Cheatham and McEver, SPE 828).

However, a poorly centralized casing or large washed-out sections can create non-uniform loading that leads to casing collapse. Ensuring good cement bonds and complete cement filling of formation-to-casing and casing-to-casing annuli are critical to achieving uniform salt loading. In addition to achieving adequate compressive strength, any cement critical to maintaining uniform loading also must have enough flexural and tensile strength to withstand the sustained casing pressure and mechanical stresses expected over the life of the well.

Salt in cement

Another important consideration in slurry design is the effect of salt on cement systems. Traditional cement designs for salt formations have used salt-saturated slurries, assuming they would bond better with salt formations, resist chemical attack, reduce the tendency for gas migration during setting and would be less likely to dissolve salt formations. However, at concentrations from about 18% by weight of water (BWOW) to saturation, salt retards thickening time, reduces compressive strength, increases thickening time, and promotes fluid loss and free-water content.

For these reasons, semi-saturated and no- or low-salt cements are appealing in many circumstances. At very low concentrations, salt actually accelerates cement setting time and improves compressive strength. However, as a low-salt slurry is pumped across a salt formation — particularly a very thick one — it will leach salt from the formation, potentially changing slurry properties dramatically.

The choice between low- and high-salt cements depends on a number of variables, including formation chemistry. For example, placing high-salt cement across a low-salt formation may allow potentially damaging osmotic pressure to build in the cement over time, as low-salt formation fluids try to penetrate the high-salt cement. Conversely, placing a low-salt cement across a high-salt formation may allow salt dissolution between the formation/cement interface and create microannuli.

Offshore Brazil

In summer 2006, an exploratory well was being drilled in about 7,000 ft (2,135 m) of water offshore Brazil. After drilling more than 10,000 ft (3,050 m), the operator reached a very thick salt formation. Previous geological studies had predicted the 6,600-ft (2,013 m) salt formation would comprise mostly slow-moving halite, with many fingers of fast-moving tachyhydrite.

The operator’s simulations showed that a heavy fluid column would provide enough hydrostatic pressure in the annulus to control tachyhydrite movement for more than 30 days — enough time to set the 14-in. casing and drill the next segment. Therefore, the operator decided to cover the tachyhydrite-laden section with a weighted drilling fluid pill and cement only the bottom few hundred feet around the 14-in. shoe.

Based on that simulation, the customer requested a design for a salt spacer and salt slurries heavier than 18 ppg. Concerns about fluid loss control resulted in an 18.5-ppg design that limited slurry salt content to 10%. Laboratory tests showed that an hour contact with the halite formation would not significantly change the slurry's properties. The spacer design had to be stable for more than 30 days at bottomhole static temperature conditions, and this was confirmed by laboratory tests.

The salt section was drilled with 11.6-ppg mud and the shoe set at 16,100 ft (4,910.5 m). After four days of logging, the open hole was considered to be stable enough to skip the weighted mud/pill. Therefore, only the 18.5-ppg spacer and cement slurry were pumped. A subsequent cement evaluation log confirmed 330 ft (100 m) of good cement, achieving the objective. The next step in the well design called for running high-strength 10-3/4-in. casing through the final halite salt section and an expected gas or oil secondary pay zone at 17,050 ft (5,200 m).

Using BJ Services’ proprietary CM FACTS modeling software (supported by compatibility testing), the recommended spacer train comprised a solvent wash, 12.5-ppg spacer and 15.8-ppg lead and gas-tight tail slurry formulations, which were pumped as designed with a good cement evaluation log result. Since that operation, The service company has successfully cemented and isolated four nearby wells using the same approach. A future well in the area will be in a high-pressure/high-temperature area with a 330 ft salt formation requiring 18-ppg salty slurries with fluid loss and gas control properties.

Gulf of Mexico

In a well on the Gulf of Mexico shelf, the deviated production string ran through a 3,500-ft (1,067.5 m) high-salt interval ending around 6,700 ft (2,043.5 m). To set the production casing, a gas-tight, salt-saturated slurry with zero free water and fluid loss under 50 ml/30 min. was required.

To meet those conditions, the service company pumped its new SaltSet cement system, which includes both gas migration control and an expansion additive to ensure a good bond. The chosen expansion agent works through a chemical reaction rather than hydrogen gas emission. The cement job also included an Ultraflush II spacer with 18% salt added after hydration of the gel. Spacer and cement systems were pumped as designed, meeting all customer expectations.

For a pending Gulf of Mexico project, the same operator is designing a deepwater well that will go through some 15,000 ft (4,575 m) of salt with high potential for damaging creep. For this deep well, BJ has designed a salt-compatible FlexSet slurry that will provide additional flexural and tensile strength aimed at ensuring the cement can withstand the temperature and pressure changes, prevent sustained casing pressure problems and protect against salt creep over the life of the well. The specially designed cement system will be used in conjunction with BJ’s patented spacer to help ensure good mud removal and prevent filtrate loss to the surrounding salt formation.

After determining accurate bottomhole circulating temperature using its proprietary software, BJ developed five FlexSet cement designs for testing. The rigorous testing is designed to determine extremely accurate Young’s modulus and Poisson’s ratios for each slurry design. Engineers then used proprietary cement stress modeling software to analyze the stresses on the cement sheath, and chose the best long-term solution (Figure 2).