Home

Some Cementing Problems

HISTORY

In 1824, Joseph Aspdin, a British bricklayer, was granted a patent for making a cement he called Portland cement after stone quarried on the Isle of Portland in Dorset, England.  However, this original Portland cement, was actually an artificial hydraulic lime similar to a material called Roman cement, a crude formulation of lime and volcanic ash used as early as 27 BCE (pozzolanic materials are still used in oil well cementing to this day).  It is sometimes reported that Aspdin did his original experiments with his kitchen oven.  The first true Portland cement is believed to have been made in Germany around 1867.

Dealing with water intrusion into oil wells, and the use of cement to prevent such movement, led to the birth of petroleum engineering.  Some primitive oil well cementing may have taken place as early as 1883, but the use of Portland cement to seal casing began in 1903 in the Lompoc Field in California.  Almond A. Perkins was the father of the two plug method of well cementing.  In 1916, Perkins employed Earl P. Halliburton, and in 1919, Halliburton set up shop on his own as the New Method Oil Well Cementing Company, changing the name to the Halliburton Oil Well Cementing Company (HOWCO) in 1920.  Perkins later sued Halliburton for patent infringement, but the case was settled with Halliburton being granted a license to use the Perkins methodology.  Halliburton eventually even licensed inventions to Perkins, but in the end Halliburton wound up buying Perkins.  Halliburton perfected the use of the measuring line and the jet mixer among other well cementing innovations.  Interestingly, over the course of the first 30 years of well cementing, waiting on cement time gradually shrank from 28 days to a mere 72 hours.

BACKGROUND

Most oil well and injection well cementing in the Illinois Basin (and other relatively shallow well areas) is done with common Portland Cement.  Portland cement is made from limestone (or other materials high in calcium carbonate) and clay or shale (iron or aluminum oxides are added if not present in sufficient amounts in the clay or shale).  These basic constituents are finely ground and mixed in the correct proportions either dry (dry process) or mixed with water (wet process).  The raw material is then fed into a rotary kiln and fired at between 2600 and 2800F causing certain chemical reactions between the raw materials.  The output of the kiln is called clinker, and is ground finely with up to about 2% gypsum to form the product we all know as Portland cement.

The American Petroleum Institute (API) sets the standards for cements used in the petroleum industry.  Here is a brief listing of the API cement designations:

Class A   Intended for use to 6,000 feet where special properties are not required.  Similar to ASTM Type I (common Portland cement).  Use 5.2 gallons of water per 94 pound sack for neat slurry*.
Class B   Intended for use to 6,000 feet where moderate sulfate resistance is required.  Similar to ASTM Type II.  Use 5.2 gallons of water per 94 pound sack for neat slurry*.
Class C   Intended for use to 6,000 feet where high early strength is needed (regular or sulfate resistant).  Similar to ASTM Type III.  Use 6.3 gallons of water per 94 pound sack for neat slurry*.
Class D   Intended for use from 6,000 to 10,000 feet (retarded).
Class E   Intended for use from 6,000 to 14,000 feet (retarded).
Class F   Intended for use from 10,000 to 16,000 feet (retarded).
Class G   Intended for use to 8,000 feet, and similar in composition to API Class B.  Use 5.0 gallons of water per 94 pound sack for neat slurry*.
Class H   Intended for use to 8,000 feet, and similar in composition to API Class B.  Use 4.3 gallons of water per 94 pound sack for neat slurry*.
*Water requirements per API for a slurry that will exhibit no water separation upon setting.

A sack of Portland cement can be hydrated with as little as 2.3 gallons of water, but the mixture cannot be pumped.  The minimum water that can be used to mix a 94 pound sack of API Class A / ASTM Type I cement, and it be pumpable, is 3.9 gallons; it will have a density of 16.85 pounds per gallon, will exert 876 psi per 1,000 feet of column, and will yield 1.0 cubic feet of slurry per sack.  The maximum water that can be used to mix a 94 pound sack of API Class A / ASTM Type I cement where there will be no settling of particulate material and no water separation is 5.5 gallons; it will have a density of 15.36 pounds per gallon, will exert 799 psi per 1,000 feet of column, and will yield 1.22 cubic feet of slurry per sack.  API recommends 5.2 gallons of water per 94 pound sack of API Class A / ASTM Type I cement; it will have a density of 15.6 pounds per gallon, will exert 811 psi per 1,000 feet of column, and will yield 1.18 cubic feet of slurry per sack.  Following the API 5.2 gallon guideline results in a requirement of 4.76 sacks of cement per 42 gallon barrel of slurry.

Garden variety Portland cement mixed with over 5.5 gallons of water per sack, and with no bentonite or other suspension additives used, will always exhibit settling.  (Note the addition of bentonite a/k/a drilling mud does not add strength to a light cement mix, but merely keeps the cement suspended until setting.  There are legitimate reasons for doing this, but not typically in shallow wells.)  Neat Portland cement mixed with 5.2 gallons of water per sack will achieve a compressive strength at 60F (typical shallow ground temperature) of about 2050 psi after 72 hours.  That same cement mixed with 10.4 gallons of water will achieve a compressive strength of a mere 425 psi under the same time and temperature conditions.  Further, the latter mix would require around an 8% bentonite addition to prevent solids settling prior to setting, or some other approved additive.

PROBLEMS

It is apparent most cement mixed in Kentucky, and in perhaps a few other jurisdictions, is being mixed with far more than 5.5 gallons of water per sack.  The biggest problem with this practice is that the resultant set cement will have a compressive strength much lower than properly mixed cement.  In the majority of these cases, no bentonite or other appropriate additive is being used, so there will be settling.  Further, the cement column weight will be lower than anticipated, and that can be a problem in wells that need the fluid column weight to kill the well for proper setting.

Wells should be plugged and completed with the highest density cement practical under the particular circumstances.  In the shallow wells here in Kentucky, that would almost always be 5.2 or at the most 5.5 gallons of water per sack neat cement.  While I have no statistics, it is likely there have been plugging failures due to improperly mixed cement.  There is no doubt there have been completion failures, but these do not show up on the regulatory radar here as much as they would in say Texas due to differences in the rules on completions (every Texas completion must have a pressure test, a bond log, and the compressive strength of cement is specified).  Notably, in overpressured wells, the heaviest practical cement slurry may be crucial to even get a successful plugging job absent setting a bridge plug.

It is customary practice to catch cement samples in little Styrofoam cups during the cement job, but many times I have seen the cementing contractors pour off the separated water before handing the sample to the oil operator or to the resident regulatory inspector (it takes a few minutes for separation to occur and it will not be evident at the mixing unit itself).  I think the problem here is a lack of recognition of the significance of separated water.  If the mix has 5.5 gallons per sack or less, there will be no free water separation (or only a trace), and the entire mass will gel / set in the sample cup (this assumes a neat cement with no additives).  The assumption seems to be if the stuff gels / sets at all it must be mixed properly.  In fact much of the settling seen in wells in the field is likely attributable to settling of slurries mixed with more than 5.5 gallons of water per sack.  And inspectors do not generally hang around long enough to really determine what the ultimate settling will be, an additional potential problem.

Ignorance of proper procedures is a large factor here, but there is a darker side to this problem.  Contractors using bulk cement equipment have the opportunity to sell the same cement more than once.  For instance, a 5.2 gallon mix will yield 1.18 cubic feet per bag, while a 10.4 gallon mix will yield 1.92 cubic feet per bag.  Since everyone uses the Halliburton cement tables in the Halliburton cementer's bible, and the tables are based on a 5.2 gallon mix, the contractor has an opportunity to charge for much more cement than is really used, and oil operators are none the wiser.  Further, contractors using centrifugal pumps, tend to mix thin to save wear and tear on equipment (and if they are using a bulk truck, they get to play the selling it twice game as well).  I want to make it clear that this is not an argument against the use of centrifugal pumps for cement mixing because I have personally mixed 15.6 pound per gallon and even a little thicker neat cement with my own centrifugal pump cement mixing rig in years gone by.  If concrete suppliers engaged in this practice, it would be noticed since it is well known that skimping on cement or using too much water results in weak concrete, but oil well cement goes down a hole and is never seen again.

This practice persists because oil operators and inspectors lack the equipment and/or knowledge to catch the problem.  A true neat slurry mixed with 5.2 gallons of water per sack of cement will have a density of 15.6 pounds per gallon (the API preferred mix).  But a slurry mixed with 10.4 gallons per sack will still weigh 13.1 pounds per gallon, and even to the trained eye looks quite heavy and thick (plain water is 8.33 pounds per gallon).  The "finger test" used by some inspectors cannot distinguish between these two examples with any degree of reliability.  Putting this example in perspective, the 5.2 gallon neat mix will exhibit compressive strength at least five times greater than the 10.4 gallon mix, and the 10.4 gallon mix will undergo some settling.  Stated bluntly, absent the use of a cement / mud balance or a hydrometer, nobody can be certain of cement density (some of the big cementing companies have continuous reading densitometers on their mixing units, but this would not likely be encountered in Kentucky).  Oil operators and regulators alike have been at the mercy of cementing contractors, and they do not always mix cement as they should.

Both the EPA Region IV Underground Injection Control (UIC) Program with jurisdiction over injection wells in Kentucky, and the Kentucky Division of Oil and Gas (DOG) with jurisdiction over oil wells, need to adopt standards that require cement to have a preferred density of no less than 15.36 pounds per gallon (this corresponds to the maximum 5.5 gallons of water per sack ratio mentioned above).  Under no circumstances should cement be pumped with a density of less than 14.0 pounds per gallon, but it is being done every day in Kentucky at present.  Inspectors should also be equipped with inexpensive drilling mud balances to make field measurements of cement slurry density.  I would hazard a guess that the majority of the cement mixed for well plugging, remediation, and new completions in Kentucky has been mixed with far too much water for optimal results.  Granted, light cements are sometimes used in deep wells, but that is seldom the case in Kentucky to date.  The major service companies like Halliburton do not play these games, but the majority of the smaller independent service companies seem to, some with a vengeance.  Why has this been allowed to go on for so many years?

See our Cement Bond Logging Overview for basic information on cement bond logging technology.  Contact us if you need more detailed information.


| Home | AnaLog Tech & Tips | PBS Tech & Tips | Cement Bond Logging | UIC |

07-17-06
Last 10-20-10