A study of 90 wells perforated with the tubing conveyed perforating system has found a correlation between underbalance pressure and formation permeability that can be used to achieve clean perforations. The data are from gas and oil producers in clean sandstones. Data for the report are from wells which were perforated, tested, acidized, and retested. There is a clear minimum underbalance line separating the data sets of wells that had clean perforations (unassisted by acidizing) from those wells that showed a significant productivity increase after acidizing. The study includes data from oil and gas wells in the Gulf of Mexico, Louisiana (Tuscaloosa), New Mexico (Morrow), Rocky Mountain Overthrust, and Alberta, Canada. Introduction Underbalance perforating or perforating the pressure in the wellbore lower than the pressure in the formation is generally acknowledged to be one of the best methods for creating open, undamaged perforations. During the few microseconds that it perforations. During the few microseconds that it takes a shaped charge perforator to create a perforation, a focused pressure wave punches a hole through the casing and into the formation. The material in the path of the pressure wave is thrust aside and the path of the pressure wave is thrust aside and the part of the formation next to the perforation may part of the formation next to the perforation may be compacted. The resultant crushing of the formation next to the perforation can reduce the initial permeability by 70% or more. Several authors have permeability by 70% or more. Several authors have noted the presence of the crush zone surrounding the perforation and have recognized that it accounts for a large part of the damage that may inhibit production. Historically, acid breakdowns were commonly used to remove this permeability damage or reduce its effect. In underbalance perforating, the pressure differential from the formation to the wellbore helps remove this crushed formation from the perforation more successfully than perforation washing or surging. The pressure differentials necessary to remove damage from a perforation is affected by pressure and flow rate. The pressure differentials necessary for perforation cleanup usually range from approximately 500 psi to over 4000 psi and have been established by trial and error in each field. This study uses information from 90 wells that were underbalance perforated, tested, acidized, and retested. The intent of the research was to determine the minimum underbalance pressure necessary to achieve undamaged perforations. It is important to optimize the amount of underbalance pressure since excessive underbalance pressure, particularly where the cement or the formation is weak, can cause the casing to collapse or the formation to disaggregate. Discussion Tubing Conveyed System One of the most popular systems for inward differential pressure perforating is the tubing conveyed system that was first described in 1975. The system involves running a perforating gun on tubing with a packer above the gun. Underbalance pressure can be achieved by swabbing or jetting out the completion fluid in the tubing to any desired height. After the packer has been set, the gun is fired either by dropping a bar, a battery pack, or by pressure firing. A perforated nipple below the pressure firing. A perforated nipple below the packer allows the formation fluids to flow into the packer allows the formation fluids to flow into the tubing after perforating. Figure 1 is a record of bottomhole pressure during underbalance perforating. The data were collected with a bottomhole pressure recorder positioned immediately above the packer. The device positioned immediately above the packer. The device communicated with the tubing through a small port.

Summary This paper details the operating implications of handling and processingnatural gas containing mercury, methods for detecting mercury, suggestedsampling procedures, and related field experiences. Introduction Mercury occurs naturally in trace quantities in air and natural gas. Although difficult to generalize, mercury concentrations in air typicallyrange1–3 between 1 ng/m3 and 10 µg/m3 (1 ppmby volume ~10,000 µg/m3). Some authors4,5 have reportedthat natural gas typically contains mercury concentrations between 1 and 200µg/m3, but concentrations are probably best evaluated on aproducing-formation basis. The implication of the effects of mercury in natural gas was not reporteduntil 1973, when a catastrophic failure of aluminum heat exchangers occurred atthe Skikda liquefied natural gas plant in Algeria.6 Investigationsdetermined that mercury corrosion caused the failure and that the mercurylikely came from an accidental source, such as test instruments used in plantand field startup. After the Skikda failure, a study of the Groningen field in Holland revealedsimilar corrosion in the gas-gathering system. CO2 was initiallythought to be the cause,7 but later investigations8pinpointed mercury, with concentrations ranging from 0.001 to as high as 180µg/m3. Phannestiel et al.8 state that most if not all of themercury in natural gas is in the elemental form and that no natural gasprocessing plant problems are suspected to have been caused by organic orinorganic mercury compounds. These statements would imply that elementalmercury is the probable cause of mercury corrosion problems. Although theconcentration of mercury in a given natural gas may be considered extremelylow, Audeh9 oberves that "its effect is cumulative as itamalgamates." Amalgamation is the formation of an alloy. To date, the most seriousproblems reported by the industry owing to mercury corrosion have been theresult of mercury forming an alloy with aluminum. Copper has also presentedsome problems.10 Saunders et al.11 observe that" brazed aluminum plate-fin heat exchangers are the predominant choice forcryogenic service. Aluminum is used due to its brazeability, excellentmechanical properties at cold temperatures, and superior heat transfercharacteristics." They further state that mercury can damage the aluminum usedin these exchangers and must be completely removed to nondetectable levels inupstream equipment. Unfortunately, complete removal is not economical. Butfortunately, the design of systems capable of minimizing operating problemsassociated with mercury in natural gas is currently being emphasized. Operating Implications of Mercury in Natural Gas The problems associated with mercury can begin at each producing well. Theinvestment required and the remote nature of wellsites prohibit mercury removalat these locations. Thus, mercury is introduced into the wellbore and gatheringsystems simply by gas production. Mercury can also be introduced into gathering systems accidentally. Testinstruments incorporating mercury and mercury-type bellows meters are oftenfound on wellsites and in gas-gathering systems. As mercury enters gathering-system pipelines, the mercury content of thenatural gas is reduced because of chemisorption onto steel pipe walls. Leeper7 suggests the following reactions as the driving force behindthis reduction:H2S+Fe2O3?FeO+S+H2O . . .(1)and Hg+S HgS. . . . (2) Trace quantities of H2S are the catalyst for the reaction ofmercury with iron oxide fiom the pipe. The mercury sulfide prccipitates and isadsorbed onto the pipe wall. Grotewold et al.10 report thatfor one 68-mile [110-km] pipeline, mercury content decreased from about 50 to20 µg/m3. This reduction is influenced by pipe-wall roughness andadhesive forces. Similar reductions are also experienced on pipe and vesselwalls in facilities and plants. Other factors influence mercury distribution in flowing streams. Mercurycondenses into the liquid phases of hydrocarbons and gas-treating chemicalsbecause of its greater solubility with higher-molecular-weight streams. Mostfacilities and processing plants that handle natural gas base have some form ofseparation, sweetening, and/or dehydration equipment. The amount of mercuryreduction resulting from contact with higher-molecular-weight solutions isdifficult to define owing to the wide range of hydrocarbon saturations andtreating conditions. Grotewold et al. report that some 50 to 60% of theinlet mercury accumulates at the bottom of the glycol absorber and that 15to20% is separated in scrubbers. This leaves 20 to 35% of the mercury enteringthe plant or facility to carry over into downstream processing equipment and/ortransportation pipelines. We have experieuced similar results at the Painter complexnitrogen-rejection-unit/natural-gas-liquid (NRU/NGL) recovery plant. Themercury concentration in the vapors off the N2 rejection column is three timeslower than that encountered on the methane liquid stream. Corrosion Mechanism of Mercury and Aluminum Elemental mercury forms an amalgam with the surface layer of the metal itcontacts. With aluminum, the amalgam is much weaker than the metal itself andis often referred to as an embrittlement. To initiate aluminum corrosion, thetightly adhering aluminum oxide layer on the surface of the aluminum must beremoved. The mercury/aluminum amalgam process removes this oxide layer. Thelayer can be removed chemically or mechanically and is catalyzed by thepresence of an aqueous electrolyte. Phannenstiel et al.8state that in their test studies, "in no instances did test results indicatethat mercury in contact with aluminum surfaces could produce gross corrosionunless condensed water was present." The mercury/aluminum amalgamationgenerally does not occur as a direct chemical reaction because the base-metalaluminum is usually protected by the oxide film. The aqueous corrosion cellforms aluminum hydroxide and gaseous hydrogen through the followingreactions:Al+Hg?AlHg . . . (3)and 2AlHg+6H2O?2Al(OH)3+3H2+2Hg. . . .(4) These reactions leave the previously amalgamated mercury free to formadditional amalgam with the base metal in a continuous corrosion process. Observations of the physical nature of the corrosion indicate that theamalgamation/corrosion is preceded by small half-moon intrusions that form acontinuous network as edges of the intrusions join together. The formation ofaluminum hydroxide can be identified by a grayish-white "whisker" appearance. Rapid pitting is common because mercury tends to collect in localizedareas.