Flow after cementing in oil and gas wells was first recognized as a problem in the 1960s but its full impact was not realized by the industry until recently. Cement slurry properties, such as density, rheology, fluid loss, gelation and setting time, have been suggested as contributing to flow after cementing One property of cement slurries that has not been discussed is water separation. An attempt has been made in this paper, based on laboratory tests and field results,
to identify the relationship between water separation in a cement slurry and the loss of hydrostatic head of a cement column and
to discover the role of designing cement slurries with little water separation to control and/or prevent flow after cementing.
The term "flow after cementing" describes the events that occur when formation fluids enter the wellbore after a string of pipe such as a casing or a liner, has been cemented in place. The intrusion of formation fluids can push the drilling mud and cement slurry from the annular space surrounding the pipe. This can often be observed at the surface as pipe. This can often be observed at the surface as a flow of fluids, usually unset cement, out of the annular space. If left unchecked, the formation fluids can channel to the surface and cause a blowout. Flow after cementing is most serious when the fluids are coming from a high pressure gas sand in a deviated hole. The problem likely occurs because the cement column does not exert the hydrostatic pressure on the formation teat would be expected based on the slurry density. Various slurry properties have been considered as factors in prevention of flow after cementing. The fluid loss characteristic of the cement has received the most attention. Fluid loss can be important when cementing across permeable sands, but as one investigator pointed out, this parameter is more likely controlled by the drilling mud filter cake and mud particle invasion than by additives in the cement slurry. The loss of hydrostatic head of the cement column can result from the thixotropic property of the unset cement. Some slurries, because of this gelation characteristic, are self supporting in the annulus. These slurries may not exert expected hydrostatic pressure at the bottom of the hole and yet have no pressure at the bottom of the hole and yet have no compressive strength. Cement slurries should have a minimum safe thickening time with a very short period between the API measured thickening time and the development of compressive strength. Other slurry properties, such as uniform mixing, rheology, and properties, such as uniform mixing, rheology, and premature setting of cement high in the annulus, all premature setting of cement high in the annulus, all have a place in controlling flow after cementing. One parameter which has not received much attention is the free water content of the cement slurry. (For a description of the standard technique for measuring free water, see the "Free Water Measurement" section of this paper.) The following Mobil field experience pointed out the effects that free water content can pointed out the effects that free water content can have on the control of flow after cementing.
Mobil Research and Development Corporation's Field Research Laboratory began working on the flow after cementing problem in 1976. Mobil Oil encountered the problem when setting the 13 3/8-inch casing in a number of wells in the High Island area of the Gulf of Mexico. Field personnel tried to solve the problem with fluid loss control. They reduced the problem with fluid loss control. They reduced the fluid loss from 1200 ml in the standard test to approximately 200 ml. The wells in which fluid loss control was tried had a more severe flow problem than previous wells. A laboratory study showed that the previous wells. A laboratory study showed that the fluid loss additive used had a detrimental effect on water separation. Slurries containing this additive had water separations in excess of 30 ml in the standard test. A slurry containing light weight cement, attapulgite clay and calcium chloride (when required as an accelerator) was recommended for these jobs.