Bachelors of Applied Science in Petroleum Engineering 2015 Year Year 3 Physical and chemical analysis of drilling fluid properties Course itle! "rilling Engineering Course Code! "#$%3001 Su&mitted to! Jasmine
Medina
Su&mitted &y! Andre' %rant (")!*51++ $a& day! 13th ,cto&er 2015 "ue date! 22nd ,cto&er 2015
1 The University of Trinidad Trinidad and Tobago, Tobago, Point Lisas Campus Esperanza Road, Road, Brechin Caste, Couva!
E-ecuti.e Summary This laboratory experiment was mainly evaluating the physical and chemical properties of drilling fluids. Six test were conducted to ascertain and correlate drilling fluid properties to their performance. As such, identifying the types of contaminants present in water based drilling fluids were of paramount importance for recommending the relevant treatments that were were app applic licabl ablee for right right type type of contam contamina inant. nt. Contam Contamina inants nts identi identifie fiedd were were calciu calcium m carbonate, oil, sodium chloride and the recommended chemical treatments were soda ash, caustic soda, gypsum and flocculation. Additionally for removal of other contaminants by mechanical means, the following treatments were subscribed Screen, forced settling and dilution. These treatments are available and are widely used in the hydrocarbon industry in order to optimi!e drilling operations while simultaneously reducing operational cost without adversely affecting the environment. environment.
" The University of Trinidad Trinidad and Tobago, Tobago, Point Lisas Campus Esperanza Road, Road, Brechin Caste, Couva!
,&/ecti.eAims! •
To determine the density and the rheological properties of original sample A and contaminated mud samples ", C, #, and $, using mud balance and viscometer
•
apparatus. To separate and measure the volumes of water, oil, and solids contained in both
•
original and contaminated %samples as stated above& via retort analysis. To ascertain the percentage of sand content of water based drilling fluids %both original and contaminated samples& by utili!ing the sand content funnel, tube, sieve'
•
mesh ( )*+m and *- hydrochloric acid solution %Cl&. To determine the filtration behaviour and wall'ca/e'building characteristics of the drilling fluid samples given, at low temperature and pressure using an A01 202T filter
•
press. To perform chemical analysis of water based drilling samples, for determination of the following o
3iltration p 4 using p strips. 5hole mud al/alinity 4 titrating with 67*8 Sulfuric acid and using
o
phenolphthalein solution as indicator. M"T and, bentonite e9uivalent 4 using 8.* m2 of methylene blue
o
solution. Calcium carbonate concentration %CaC$ :& 4using ;m2 of .86
o
and ;8 =pm versanate hardness titrating solution. Calcium concentration 4using
o
hardness titrant solution. Chloride ion content 4using 8.8;6 %67*8& sulphuric acid, potassium chromate indicator solution %> ;Cr$?&, and 8.8;@;6 Silver 6itrate
•
Solution %Ag6$ :&. Sodium Chloride and potassium chloride content using phenolphthalein indicator solution, 8.8;6 Sulphuric acid solution, potassium chromate indicator solution, 8.;@;6 silver nitrate solution, standard sodium perchlorate solution and a hand cran/ centrifuge.
heory! Bacground information!
The successful completion of an oil well and its cost depends on considerably on the extent of the properties of the drilling fluid. Many re9uirements are placed on the drilling # The University of Trinidad and Tobago, Point Lisas Campus Esperanza Road, Brechin Caste, Couva!
fluid. 1n the past, main purpose of the drilling fluid was to serve as a vehicle to remove cutting from the well bore, however in recent times, the applications of drilling fluids has been more diversified %ray, Caenn and #arley B@:&. ence in rotary drilling, the principal functions performed by the drilling fluid includes the following Carry cuttings from beneath the bit, transport them up the annulus, and permit their •
•
separation at the surface. Cool and clean the bit. educe friction between the drilling string and the sides of the hole. Maintain stability of uncased sections of the borehole. 0revent the inflow of fluids 4 oil, gas, or water 4 from permeable roc/s that were
•
penetrated. 3orm a thin and relatively impermeable filter ca/e which seals pores and other
• • •
•
openings in formations penetrated by the bit. Assist in the collection and interpretation of information available from drill cuttings, cores, and electrical logs.
#rilling fluids are categori!ed in accordance to their base .i.e. water based and oil based muds. 5ater based muds are consist of solid particles suspended in water or brine. 1n some cases, oil may be emulsified in water, in these cases water is considered as the continuous phase. 5hereas, oil based muds comprise of solid particles suspended in oil. 1f water or brine is emulsified in oil then the oil is considered to be the continuous phase. Another type of drilling fluid is gas. This is where drill cuttings are removed by a high velocity stream of air or natural gas. 3oaming agents are added to remove minor inflows of water. 1n water based muds, the solids consist of clays and organic colloids added to provide the re9uired viscous and filtration characteristics, heavy minerals %generally barite are added to increase density when needed& and solids from the formation that become dispersed in the mud in the course of drilling. The water contains dissolved salts either from contamination with formation waters or purposely added for any number of reasons. The following sections are brief introductions of the six diagnostic tests that will be performed on the drilling fluid samples. "ensity and #heological properties!
The density #rilling fluid must be maintained to provide the re9uired hydrostatic head to prevent flux of formation fluids, but not so high as to cause loss of circulation or unfavourably affect the rate of drilling and formation damage. Conse9uently, one of the first test to be performed on a drilling rig is mud weight or density. $ The University of Trinidad and Tobago, Point Lisas Campus Esperanza Road, Brechin Caste, Couva!
3igure showing typical diagram of mud balance %"ourgoyne Jr., et al. B@D& 1n this experiment, the density and rheological properties of the original and contaminated samples is being performed. The apparatus used to conduct this test was the mud balance %shown in figure above&. The test consists of essentially of filling the cup with a mud sample and determining the rider position re9uired for balance. The balance is calibrated by adding lead shot to a calibration chamber at the end of the scale. 5ater usually is used for the calibration fluid. The density of fresh water is @.:: lbm7gal. The drilling fluid is normally degassed before being placed in the mud balance to ensure an accurate measurement %"ourgoyne Jr., et al. B@D&. 1n this section of the experiment, rheological properties of the drilling samples will be measured using the rotational viscometer.
other rotator speeds, the apparent viscosity is given by
=
300θ N N
, where E 6, is the
dial reading in degrees and 6 is the rotor speed in revolutions per minutes. The viscometer could also determine rheological parameters that exhibit non'6ewtonian fluid
% The University of Trinidad and Tobago, Point Lisas Campus Esperanza Road, Brechin Caste, Couva!
behaviour for example, the flow parameters of "ingham plastic model as shown in figure ; below.
3igure ; showing 6ewtonian and 6on'6ewtonian curves %>ing 3ahd Fniversity of 0etroleum and Minerals, ;88:&
#etort Analysis!
3igure : above shows the retort distillation apparatus consisting of three principal components a heating unit, a condenser and a receiver. The heating unit, is used to & The University of Trinidad and Tobago, Point Lisas Campus Esperanza Road, Brechin Caste, Couva!
bombard the reservoir roc/ sample with extreme heat. oc/ samples can either be crushed or small cylindrical core plugs in dimensions... These roc/ samples, either consolidated or non'consolidated, are generally weighed before placing them in the retort. eat is dispensed at either in stages or directly to temperatures as high as D*8 8 C resulting in the vapori!ation of oil, and water. This vapori!ed oil and water, is then condensed in the condenser and collected in a small receiving graduated cylinder, where the volumes of oil and water can be measured directly. 6o further extraction of pore fluids %>, ;88D& are indicated by the presence of a hori!ontal plateau in the plot of collected oil and water volume vs. the heating times. Sand content of 'ater &ased drilling fluids!
According to "aroid 1ncorporated %;8*&, measurement of the sand content of mud should be made regularly, because excessive sand ma/es a thic/er filter ca/e, this in turn causes abrasive wear of pump parts, bit and pipe, may also settle when circulation is stopped and interfere with pipe movement or settling of casing. Sand content %A01& method is defined as the percentage by volume of solids in the mud that are retained on a ;88'mesh sieve. "elow shows a table that defines and characteri!e sieve si!es for different types of sand.
' The University of Trinidad and Tobago, Point Lisas Campus Esperanza Road, Brechin Caste, Couva!
3igure ? below shows the standard A01 sand sieve that will be used for determination of sand content in water based drilling fluids %ray, Caenn and #arley B@:&.
AP( fluid loss!
3luid loss is usually termed as the loss of a mud filtrate %li9uid phase& into a permeable formation that is being drilled. "ecause of positive differential pressure %i.e. the pressure difference between the mud pressure in the wellbore and the formation pore pressure&, the mud filtrate tends to flow into the formationG Conse9uently, this creates a an accumulation of mud solids deposited on the wellbore walls, thus forming what is generally referred to as mud ca/e %filter ca/e&. 3urthermore, initial loss of filtrate to the formation at time !ero is termed as initial spurt loss. After a mud ca/e is formed, the presence of any loss of filtrate is categori!ed as the continuous loss %A!ar H Samuel, ;88)&. 1n the hydrocarbon industry, there are two types of filtration involved in drilling an oil well static filtration and dynamic filtration. Static filtration occurs when the mud is being not being circulated and filter ca/e growth is undisturbed. owever, dynamic filtration occurs when the mud is circulated and growth of the filter ca/e is limited by the erosive action of mud stream. The filtration of properties of drilling fluids are generally evaluated and controlled by the A01 filter loss test which is a static test. owever, because a static test being performed, this is not a reliable guide to the downhole filtration which is usually dynamic %ray, Caenn and #arley B@:&. ( The University of Trinidad and Tobago, Point Lisas Campus Esperanza Road, Brechin Caste, Couva!
Chemical Analysis
1n order to determine the concentration of various ions present in drilling fluids, a wide array of chemical analyses will be performed. These tests include determination for $ ', Cl', and Ca;I, which are re9uired to complete the A01 drilling mud report form. 3urthermore, a titration apparatus is used to conduct these type of tests. Titration involves the reaction of a /nown volume and concentration. The concentration of ion to be tested will be determined from /nowledge of the chemical ta/ing place %"ourgoyne Jr., et al. B@D&. =xperience has shown that certain chemical analyses are useful in the control of mud performance, for example, an increase in chloride content may adversely affect the mud properties unless the mud has been designed to withstand contamination by salt. Those analyses that have been found to be adaptable to use in the field have been included in A01 " : " %ray, Caenn and #arley B@:&. Salt Analysis4 determination of sodium and potassium chloride content
A sample of mud filtrate %neutrali!ed, if al/aline& is titrated silver nitrate solution, using potassium chromate as indicator. The results are usually reported in parts per million chloride ion, although actually measured in terms of mg Cl ' ion per 888cm: of filtrate. 1n order to determine the chloride content of an oil mud, the sample will be diluted with a mixture of =xosol and isopropyl alcohol %:& and diluted with water, neutrali!ed to the phenolphthalein end point and then titrated the usual way %ray, Caenn and #arley B@:&. owever, in this laboratory session only the salinity of water based mud samples will be examined.
Procedure! As per la& manual
) The University of Trinidad and Tobago, Point Lisas Campus Esperanza Road, Brechin Caste, Couva!
#esults!
1* The University of Trinidad and Tobago, Point Lisas Campus Esperanza Road, Brechin Caste, Couva!
a&le 1 a&o.e sho'ing results for si- e-periments performed on drilling fluid samples A4"
11 The University of Trinidad and Tobago, Point Lisas Campus Esperanza Road, Brechin Caste, Couva!
Graph 1 showing spurt loss V vs √t 0ampe B
Linear +0ampe B-
0ampe C
Linear +0ampe C-
rigina 0ampe
Linear +rigina 0ampe-
0ampe 2
Linear +0ampe 2-
0ampe 3
Linear +0ampe 3-
f+- . "!*) / *!&(
f+- . 1!'$ / *!
%$f+- . 1!%( / *!1#
f+- . 1!"' / *!#1
f+- . 1!1# / *!1(
1" The University of Trinidad and Tobago, Point Lisas Campus Esperanza Road, Brechin Caste, Couva!
Calculations!
The following are sample calculations for each experiment conducted. 6sing data from original sample for all sample calculations7 e-cept 'here specified "ensity and #heological Properties
0< %0lastic viscosity, %lbs788ft ;&7:88rpm& E D88'E:88 K0 %yield point in lbs788ft ;& E:88 4 0< PV = θ 600
− θ 300 = 40 − 25 = 15lbs / 100 ft 2 / 300rpm YP = θ 300 − PV = 25 − 15 = 10lbs / 100 ft 2
#etort Analysis
100 (Oil volume collected, mL) Sample Volume, mL
= V w =
⇒ Vo = ⇒ Vw =
Vs
=
100 (Water vol ume collected, mL) Sample Volume, mL 100 − (Vo
)
+ Vw
100(0)
= 0% 10 100(0!)
= % 10 ⇒ Vs = 100 − (0 + ) = 1% Chemical Analysis
Methylene "lue Capacity %M"T& methylene "lue, m27#rilling fluid, m2 "entonite e9uivalent, lb7bbl * %Methylene "lue, m2&7#rilling 3luid, m2 "ml =4 2!0 ml ⇒ 5( 4) / 0!#1 = 2"lb / bbl
⇒ MBT =
1# The University of Trinidad and Tobago, Point Lisas Campus Esperanza Road, Brechin Caste, Couva!
Total hardness for sample B as calcium, mg71 ?88 x %< =#TA7
400 $ 1cm 3 1cm 3
= 400
6aCl' #etermination cClN 8888 x %
∴ NaCl = 1!65 $ 12100 = 1,650 mg / I Conversion of Mg71 to weight percent and 00M at D@ 8 3 gives Q@8,888 ppm. Fsing the graph of 6aCl weight- against 6aCl mg71 and @8,888ppm gives * wt- of 6aCl. >Cl determination c>ClN, ppb %)7
ClN, ppb %)78.D*& x .D ;?.B; ppb c>INI *88 x c>ClN, ppb c>INI *88 x ;?.B; @),:@?.D y ml >Cl ppt 8.8:B: x %lb7bbl ppt & I 8.;8?; y ml >Cl ppt 8.:B: x .D I 8.;8?; ?.)D: lb7bbl
1$ The University of Trinidad and Tobago, Point Lisas Campus Esperanza Road, Brechin Caste, Couva!
"iscussion! "ensity and #heological properties!
#ensity or mud weight, was determined by weighing a precise volume of mud and dividing it by the volume. The mud balance was the instrument utili!ed to obtain the density for both original and contaminated mud samples. To date in the petroleum industry, the mud balance provides the most convenient way of obtaining a precise volume. The procedure that is normal used on a drilling rig, is to fill the cup with mud, put on the lid, wipe off the excess mud from the lid, move the rider along the arm until a balance is achieved and the density was read at the side of the rider towards the /nife edge %ray, Caenn and #arley B@:&. #ensity could be expressed in pounds per gallon %lb7gal&, pounds per cubic foot %lb7ft :&, and grams per cubic centimetre %g7cm :& or as a gradient exerted per unit depth. As shown in table one, the density for the original sample was recorded at B.8 ppg, whereas samples A to $ densities were recorded at ).), B.;*, @, B.; and B.;* lb7bbl respectively. The disparity in densities between the original mud sample and the contaminated mud samples could be attributed to the following •
Contaminated samples may have excess A01 barite. A01 barite is a dense, inert mineral having a specific gravity of approximately ?.; and this could be added to any
•
clay 7 water mixture to increase density %"ourgoyne Jr., et al. B@D&. 3urthermore, the contaminated samples could also contain inert solids. These solids are termed inert, because they do not hydrate with other components of the mud. 1nert solids are generally classified as sand, silt, limestone, feldspar and also A01 barite. 1n this experiment mud samples ", # and $ have higher densities than the original mud sample. 1t was observed from table one above, that these samples have relatively more percentage sand content than the original sample. This observation was supported by the presence of calcium carbonate in samples ", # and $. 5hen these samples were in the same mixture containing hydrogen chloride, they
1% The University of Trinidad and Tobago, Point Lisas Campus Esperanza Road, Brechin Caste, Couva!
effervescence and their sand content after acidification was reduced. This reaction was not observed in the original mud sample. 5hen inert solids such as sand are present in drilling fluids, they adversely affect the functionality of the drilling fluidG such as, they may increase the frictional pressure drop, in the fluid system, but they do not greatly increase the ability to carry the roc/ cuttings to the surface. The filter ca/e formed from these solids is thic/ and permeable rather than thin and relatively impermeable. Conse9uently, delay in drilling activities arises dud to stuc/ pipe, excessive pipe tor9ue and drag, loss of circulation and poor cement bonding to the formation %"ourgoyne Jr., et al. B@D&. 1n addition excessive mud density due to inert solids could possibility increase the hydrostatic pressure on the borehole walls so much so that the hole fails in tension. This failure is /nown as induced fracturing. This phenomenon is where mud is lost into the facture that formed and the level of the annulus falls until e9uilibrium conditions are obtained. Another disadvantage of excessive mud densities is their adverse influence on rate of penetration. This occurrence have been proven by laboratory experiments and field experience that in the event of mud overbalance especially drilling in very low permeability roc/s, the rate of penetration is significantly reduced. Also, a high overbalance pressure increases the probable ris/ of stic/ing the drill pipe. 3inally, these aforementioned problems that could arise from the presence of inert solids in drilling fluids causes unnecessary drilling costs and overruns. To date, excess concentration of inert solids in drilling muds can be reduced to a desirable levels by screening, forced settling, chemical flocculation and dilution. #heological properties
The rotational viscometer was used to measure the rheological characteristics of the mud samples prepared. The mud was sheared at a constant rate between an inner bob and an outer rotating sleeve. The viscometer, was also used to determine rheological parameters that described 6on'6ewtonian fluid behaviour. Two flow parameters that were re9uired to characteri!e the mud samples that follow the "ingham plastic model were plastic viscosity and yield point. The plastic viscosity, c0, in centipoise was computed using c0 E
'E:88
D88
where ED88 was the dial reading with the viscometer operating at D88 rpm and E :88 was e9ual to the dial reading with the viscometer operating at :88 rpm %"ourgoyne Jr., et al. B@D&. The shear stress divided by the shear rate %at any given rate of shear& is /nown as the effective or apparent viscosity. =ffective viscosity decreases with the increase of shear rate, and was 1& The University of Trinidad and Tobago, Point Lisas Campus Esperanza Road, Brechin Caste, Couva!
therefore a valid parameter for hydraulic calculations only at the shear rate at which it was measured. Moreover, the decrease in effective viscosity with increase in shear rate is /nown as shear thinning, and normally this is a desirable property, because of the effective viscosity would be relatively low at the high shear rates prevailing in the drill pipe, thereby reducing pumping pressures, and relatively high at low share rates prevailing the annulus, thereby increasing cutting carrying capacity. The fact that the consistency curves %illustrated in graph & of clay muds intercept at the stress axis %i.e. y'axis&, at a greater value that !ero was indicative of gel structure development. Clay particles in drilling fluids are highly anisodimensional and can build a structure at very low solid concentrations, because of interaction between attractive and repulsive forces. At low shear rates the behaviour of clay particles was influenced by these forces and as a result, the particles viscosity were relatively high, but as shear rate increases, the particles gradually align themselves in the direction of flow and the viscosity then becomes largely dependent on the concentration of all solids present in the mud. This phenomena was observed for all drilling sample fluidsG because of these occurrences, the degree of deviation from linearity in the "ingham plastic consistency curves %as shown in graph one above& of drilling muds differs from mud to mud in the rotary viscometer and this was depended on particle si!e and shape, and concentration of bentonite %ray, Caenn and #arley B@:&. This phenomenon directly affected the filter ca/e properties developed by the samples in the lab as seen in table. This type of behaviour was observed with samples with low solid muds containing a high proportion of clay particles and high solid muds such as barite. Fnfortunately, it is highly challenging to determine the linearity of the consistency curves, other than by measurement in a multispeed rotary viscometer. 1n practice the most widely use of the 0< and K0 9uantities is for the evaluation of drilling mud performance and is used as a guide for drilling mud treatments. Thus 0< is sensitive to the concentration of solids and this is indicative of dilution re9uirementsG K0 is sensitive to the electrochemical environment, and hence indicates the need for chemical treatment %ray, Caenn and #arley B@:&. Fsually, the consistency curve of a "ingham plastic in a rotary viscometer should be linear at rotor speeds above that re9uired to /eep all the fluid in the annulus in laminar flow. 1n reality however, drilling fluids are not ideal "ingham plastics and as such they deviate from linearity at low shear rates.
1' The University of Trinidad and Tobago, Point Lisas Campus Esperanza Road, Brechin Caste, Couva!
A third non'6ewtonian rheological parameter called gel strength, in units of lbf788 s9 ft ; was obtained by noting the maximum deflection when the rotational viscometer was turned on at a low rotator speed of : rpm. el strength was termed as observing the maximum deflection before the gel brea/s. el strength for all the samples were measured after allowing the mud to stand 9uiescent for 8 seconds, the maximum dial deflection obtained when the viscometer was turned on was the initial gel strength. The gel strength of fresh water clay muds, increases with time after agitation has ceased, a phenomenon called thixotropy. 3urthermore, after standing 9uiescent the mud was subRected to a constant rate of shear, its viscosity decreases with time as its gel structure was bro/en up, until an e9uilibrium viscosity was reached. Thus the effective viscosity of a thixotropic mud was time dependent as well as shear'dependent.
#etort Analysis
The mud samples were placed in a steel container and were heated %approximately *D I7' ;; 8 C& until the li9uid was vapori!ed. The vapours passed through the condenser and were collected in a graduated cylinder. The volumes of the respected samples were measured and then converted to a percentage based on the volume of whole mud in the retort cup.
3rom e9uation one above, the solids both suspended and dissolved, were e9ual to 88minus the li9uid percent. This procedure also gave the percent of oil in the mud sample. Conse9uently, it was found that sample C was the only sample contaminated with oil. The mud samples were then subRected to further tests to elucidate the nature of contaminated oil content. The retort procedure was a very rapid and simple techni9ue, however, the retort distillation method has notable disadvantages. 3irstly, the roc/ samples were completely destroyed and secondly, high temperatures were re9uired. owever, the application of extreme heat was unavoidable because, oil in the reservoir roc/ samples contained very high molecular weight or high boiling point substituents. Conse9uently, the application of very high temperatures was essential to ensure that all the oil was completely extracted from the roc/ samples %>, ;88D&. Fsing elevated temperatures of this magnitude resulted in the following errors
1( The University of Trinidad and Tobago, Point Lisas Campus Esperanza Road, Brechin Caste, Couva!
At such high temperatures, the water of crystalli!ation within the roc/ was driven off, causing the water recovery values to be greater than the pore water %>, ;88D&. igh temperatures also may fracture and the co/e in the oil causing the collected oil volume not to correspond to the volume of oil initially in the roc/ sample. The crac/ing and co/ing of the hydrocarbon molecules, may li/ely to reduce the li9uid volume and also in some cases may also coat the internal walls of the roc/ sample itself. The water of crystalli!ation and the crac/ing and co/ing of hydrocarbons was 9uantified in =mdahl based on the core analysis of 5ilcox sands in which fluid saturations were measured by the retort distillation method, indicating an error of around ::- in the water saturation with the volume of oil recovered and the volume of oil in the sample varied due to < oil actually in the sample .;B@ %<
8.@*B
oil collected in receiver&..........................;. =9uation ; indicates that the volume of oil recovered or collected in the receiver was decreased due to crac/ing and co/ing of the hydrocarbon molecules %>, ;88D&. 1n addition to these errors, other practical errors could also occur in the retort distillation method, such as formation of oil'water emulsions that do not allow accurate volume measurements and the absence of clear demarcation between the plateaus of pore space water and the water of crystalli!ation which could introduce uncertain measurement of water volume. Sand content and 'ater &ased drilling fluids
The sand content test was a measure of the amount of particles larger than ;88 mesh present in mud samples. =ffectively this test defines the si!e and not the composition of the particles %ray, Caenn and #arley B@:&. The mud samples were first subRect to dilution by adding mud and water to the respective mar/s inscribed in the glass tube. The mixture was then sha/en and poured through the screen in the upper of cylinder, and then washed with tap water until clean. The substance that remain on the screen was then bac/washed through the funnel into the glass tube and allowed to settle and finally the gross volume was read from the gradulations on the bottom of the tube. As mentioned earlier, the presence of excessive sands in drilling fluids have direct catastrophic problems on drilling operations of a well and as such there were four methods that could be employed to prevent a high concentration of inert solids. These were •
Screening. This method is usually applied first in processing the annular mud stream. This allows the removal of most of the solids before their si!e has been reduced to the si!e of the A01 barite particles. A01 specifications for commercial barium sulphate 1)
The University of Trinidad and Tobago, Point Lisas Campus Esperanza Road, Brechin Caste, Couva!
re9uire that B)- of the particles pass through a ;88'mesh screen. 0articles less than approximately )?+m in diameter will normally pass through the ;88'mesh screen. 3orced Settling. 5hen natural settling failed to screen out inert particles, devices
•
such as hydroclones and centrifuges are utili!ed to increase the gravitational force acting on the particles. At present, both devices are used as forced settling instruments with unweighted muds %"ourgoyne Jr., et al. B@D&. Chemical flocculation. The removal of fine active clay particles could also be used by
•
adding chemicals that cause the clay particles to flocculate or agglomerate into larger units. $nce agglomeration of fine clay particles have been achieved, separation can be facilitated more easily. #ilution. This method re9uires discarding a portion of additives used in previous mud
•
treatments.
AP( fluid loss
The A01 fluid loss test was used to determine the static filtration characteristics of the mud and the need for treatment with fluid loss additives %only used for water based muds&. The filter press was used to determine, the filtration rate through a standard filter paper and the rate at which the mud'ca/e thic/ness increases on the standard filter paper under standard test conditions. This test was indicative of the rate at which permeable formations were sealed by the deposition of a mud'ca/e after being penetrated by the bit. 1f unit volume of a stable suspension of solids was filtered against a permeable substrate, and x'volumes of filtrate were expressed, then 'x volumes of ca/e %solids plus li9uid& would be deposited on the substrate. Therefore if c be the volume of the ca/e and w the volume of the filtrate Qc Qw
=
1 − x x
and the ca/e thic/ness %h& per unit area of ca/e in unit time would be h
=
1 − x x
$ Qw
"* The University of Trinidad and Tobago, Point Lisas Campus Esperanza Road, Brechin Caste, Couva!
owever, #arcys law stated that dq dt
= KP µ h
Therefore dq dt
=
KP µ Qw
$
x 1 − x
1ntegrating Qw 2
=2
KP
$
µ
x 1 − x
$ t
Then substituting Qw
Qw
2
2
=
=
2 KP Qw $ $ t µ Qc
2 KPA 2 µ
$
Qw Qc
1f the area of the filter ca/e was A, then $ t
This is the fundamental e9uation governing filtration under static
environment. According to "ourgoyne et al %B@D& the filtrate volume should be proportional to the s9uare root of the time period used. Thus, the filtrate collected after ).* minutes should approximately be half the filtrate collected in :8 minutes. This phenomenon was observed for all drilling fluid mud samples. . 1n order to determine if a significant spurt loss of volume of filtrate was observed for each of the mud samples the volume of filtrate collected vs s9uare root of time %Ut& was plotted on a graph %see graph above&. The spurt loss was determined by extrapolation and the following e9uation was utili!ed < :8 ;%<).*'
observation that both original sample and sample " were more permeable and thus a higher fluid loss relatively to the other mud samples. Chemical Analysis p8 paper strip method
The p, or hydrogen ion concentration, was a measure of the relative acidity or al/alinity. The p values ranges from 8 to ?, with 8'D being acid, ) being neutral and @'? being al/aline. 3or the purpose of this experiment, p strips were used. These strips change colour in accordance with the acidity or al/alinity of the filtrate or mud. The p determined for the original sample and contaminated samples A to # were @, 8,8,B, and 8 respectively. The p of mud plays a maRor role in controlling the solubility of calcium. At high p values 4as shown above', calcium solubility was very limitedG this ma/es high p mud suitable for use in the drilling of carbonate formations, which normally were susceptible to erosion and dissolution by freshwater mud. The p value was also an important indicator for the control of corrosion. According to A!ar and Samuel %B@?&, a minimum of B.* should always be maintained to prevent oxygen corrosion of casting, drill pipe, etc. A high p tends to disperse the active clays in the mud.
9hole mud alalinity (P m )
Al/alinity refers to the ability of a solution or mixture to react with an acid. The phenolphthalein al/alinity refers to the amount of acid re9uired to reduce the p to @.:, the phenolphthalein end point. The phenolphthalein al/alinity of the mud and mud filtrate is called the 0 m and 0f , respectively. The 0 f test includes the effect of only dissolved bases and salts while the 0 m test included the effect of both dissolved and suspended bases and salts. The methyl orange al/alinity refers to the amount of acid re9uired to reduce the p to ?.:, the methyl orange endpoint. The methyl orange al/alinity of the mud and mud filtrate is called the M m and Mf , respectively. The A01 diagnostic test include the determination of 0m, 0f and Mf. The 0f and Mf test were designed to establish the concentration of hydroxyl, bicarbonate, and carbonate ions in the a9ueous phase of the mud. At a p of @.:, the conversion of hydroxides to water and carbonates to bicarbonates was essentially complete %"ourgoyne Jr, Chenevert, Millheim, H Koung Jr, B@?&. The bicarbonates originally present in solution do not enter the reactions. Thus, at a p of @.:, $' I I
$, and C$:;' I I
C$':'. ""
The University of Trinidad and Tobago, Point Lisas Campus Esperanza Road, Brechin Caste, Couva!
As the p was further reduced to ?.:, the acid then reacts with the bicarbonate ions to form carbon dioxide and water C$:' I I
C$; I $.
owever, one disadvantage of this type of test is that in many mud filtrates, other ions and organic acids are normally present that can adversely affect the Mf test. The 0f and 0m test results indicate the reserve al/alinity of the suspended solids. As the $'N solution was reduced, the lime and limestone suspended in the mud would go into solution and tend to stabili!e the p. This reserve al/alinity generally was expressed as an e9uivalent lime concentration. Converting the Ca%$&; concentration from 8.8;6 to field units of lbm7bbl yields 8.8; gew72 x :).8*7g72 8.;D lbm7bbl. Thus, free lime was by 8.;D %0m 4fw L 0f&, where fw was the volume fraction of water in mud which was reported to be 8.88B@. :B and Bentonite e;ui.alent
This test gives an estimate of the cation exchange capacity of mud solids as well as to indicate the amount of active clays in the mud system. Also this test could be used to determine colloidal characteristics of clay minerals. A standardi!ed solution of methylene blue dye was added to ml of mud that has been treated with hydrogen peroxide and sulfuric acid and was then gently boiled to decompose the polymers and organics %which have a very high exchange capacity and would otherwise interfere with the test&. The methylene blue was added in 8.* ml increments until the mud solids no longer absorb the dye. This endpoint was determined by putting a drop of the solution on a standard 5hatman filter paper. 5hen the dye was in excess, a halo of free dye formed around the blue dot. The halo that formed was tur9uoise blue in colour and was very distinct form the blue colour of the dye. This was reported as e9uivalent lbs7bbl bentonite. Calcium Car&onate "etermination
After retort and sand analysis were performed on both original and contaminated samples, determination of calcium carbonate content of the water based drilling samples were then carried out. 1n order to determine the calcium content of both samples, the total hardness of the samples were estimated using the versanate method. The hardness of water or drilling fluid was due to mainly the presence of calcium and magnesium ions. 5hen =#TA was added to water, it combined with calcium ions and the endpoint was determined in the presence of calver indicator. 5hen all the calcium ions was complexed with the =#TA "# The University of Trinidad and Tobago, Point Lisas Campus Esperanza Road, Brechin Caste, Couva!
solution, it gave a colour change at a p of ;':. The colour change observed in the solution was from a wine colour to blue blac/. 1t was observed that sample # showed presence of calcium carbonate contaminant. 1n the petroleum industry, the practice of chemical removal of contaminants are utili!ed. The addition of chemical contaminants to the drilling fluid, either at the surface or through the wellbore, produces an imbalance in the chemical e9uilibrium of the fluid, which can cause serious rheological or drilling problems to develop. 3or example, when calcium enters the mud, sodium montmorillonite will convert to calcium montmorillonite, which first produces flocculation and eventually aggregation of the montmorillonite. This is often desirable to remove the calcium by chemical treatment. 1n most cases, calcium is removed from the mud system by adding soda ash %6a ;C8:& which forms in soluble calcium carbonate ↓
⇒
Ca;I I ;$' I 6a;$:
CaC$:
I ;6aI I ;$'
3urthermore, if cement or lime get into the mud, the p usually increases to unacceptable levels because of the hydroxyl ions as well as calcium have been added. 1n these circumstances, either sodium acid pyrophosphate SA00'6a ;;0;$) or sodium bicarbonate is usually added. 5hen SA00 is added the following occurs ↓
→
Ca;I I ?$' I 6a ;;0;$)
Ca;0;$)
I ;6aI I;$' I ;;$. 1n this reaction, calcium
was removed and the four hydroxyl ions on the left side of the e9uation are reduced to two hydroxyl ions on the right side %"ourgoyne Jr, et al. B@?&. Calcium test9ater 8ardness
The mud hardness indicates the amount of calcium suspended in the mud as well as the calcium in solution. This test usually is made on ypsum'treated muds to indicate the amount of excess CaS$ ? present in suspension. A small contaminated sample of mud was first diluted to *8 times its original volume with deioni!ed water so that any undissolved calcium or magnesium compounds can go into solution. Since the mud samples were diluted *8 times their original volume, a *8 cm : sample was titrated to determine the calcium and magnesium present in cm : of mud. 5ater containing large amounts of Ca ;I and Mg ;I ions is /nown as hard water. These contaminants were often present in the water available for use in the drilling fluid. 1n addition, Ca ;I can enter the mud when anhydrite %CaS$ ?& or ypsum %CaS$?.;;$& formations are drilled. Cement also contains calcium and can contaminate the mud. The total Ca
;I
and Mg ;I concentration was determined by titrating
with a standard %8.8; 6&
contains sodium
;I
and Mg ;I . The chelate ring structure very stable and essentially removes the Ca ;I and Mg ;I from solution. #isodium ethylenediaminetetraacetic acid %=#TA& plus calcium yields the =#TA chelate ring See chemical reaction below %"ourgoyne Jr., Chenevert, Millheim, H Koung Jr., B@?&.
Magnesium ion forms a wine red complex with the dye =riochrome "lac/ T. Since the solution containing both Ca
;I
and Mg
;I
was titrated in the presence of this dye, the
Mg I ; 6a$
Mg%$&;I ;6a I
Chloride (on content
Salt can enter and contaminate the mud system when salt formations were and when saline formation water enters the well bore. The chloride concentration was determined by titration with silver nitrate solution. This caused the chloride to be removed from the solution as AgCl, a white precipitate Ag I I Cl'
AgCl
"% The University of Trinidad and Tobago, Point Lisas Campus Esperanza Road, Brechin Caste, Couva!
The endpoint was detected using a potassium chromate indicator. The excess AgI present after all Cl' has been removed from the solution reacts with the chromate to form Ag;Cr$?, and orange'red precipitate ; AgI I Cr$? Ag;Cr$?
Since AgCl was less soluble than Ag ;Cr$? , the latter cannot form permanently in the mixture until the precipitation of AgCl has reduced the Cl'N to a very small value. 3or titration, .8; 6 Ag6$: concentration was used %"ourgoyne Jr., Chenevert, Millheim, H Koung Jr., B@?& Salt analysis
)ml of filtrate was measured and :ml of standard sodium perchlorate solution was added to this. The resultant mixture was then centrifuged at @88 rpm for one minute and the precipitate volume were recorded which was 8.D* and this was extrapolated on the calibration curve to obtain lb7bbl of >Cl. The maximum density of a solids'free fluid depends on the type of salt used. =ach salt has a maximum concentration before it reaches saturation. The table below indicates the maximum densities of various brines. Thermal expansion of the water affects the density of clear brine. At elevated temperatures the density decreases. #ensities were reported at a specific temperature such as )8 83. Combinations of salts can be used to economically achieve densities from @.:? to B.; ppg %eo #rilling 3luids 1nc, ;8?&.
V?7:; and they were both thin soft and pliable. owever, sample As mud thic/ness was recorded to be @7:; which was twice the mud thic/ness of the original mud sample. This could be attributed to contaminants such as 6aCl and hardness of water. ardness of water means that there were calcium and magnesium ions present in the mud system thereby reducing sodium montmorillonite to expand and hydrate in water. 3urthermore, high concentration of salts in water could greatly affect the ability of
2
some clays to hydrate in water. 0ercent solids in original sample was found to be -G in contrast to the contaminated
samples " and C contained ?- bentonite and D- oil. 3 Sample A has sand and Sample # has carbonates. A01 3luid loss uestions "& The University of Trinidad and Tobago, Point Lisas Campus Esperanza Road, Brechin Caste, Couva!
. The
original sample has more concentration of bentonite which acts as a
viscosifier and readily hydrates in water thus increasing viscosity of the mud and
;. :. ?.
decreasing fluid loss. emoval of contaminants and adding other types of high yield clays such as smectite or attapulgite as well as CMCs and other polymers. Some factors are 3iltrate viscosity, ca/e permeability, pressure differential. See graph. Spurt loss is calculated by the following a. The spurt loss of the cell can be obtained by extrapolation to !ero time and finding the gradient. 3or the original sample Time
3iltrate
?.?); cm:
).*
8
10 − 4!4#2 $ 4!4#2 − #!5 − 1 Qw
2
*. Vf
=
=
2 KPA 2 µ
$
Qw Qc
1 = 1!2cm
3
$ t
fsc − 1 $ A fsm
2 k ∆ p
t µ
This e9uation indicates that the filtrate volume is
proportional to the s9uare root of the time period used. Thus, the filtrate collected after ).* minutes should be half the filtrate volume in collected after :8 minutes. 1t was concluded that filtration rate increases with temperature because the viscosity of the filtrate is reduced.
#etermination of >Cl concentration 9uestions . Sample A had 6aCl. Conclusion!
"' The University of Trinidad and Tobago, Point Lisas Campus Esperanza Road, Brechin Caste, Couva!
#rilling fluid densities for water based 2ignosulfonate mud was obtained using a non' pressured mud balance. $ne of the product formulation was "entonite, which was added for mud viscosity, gel strength and even fluid loss control. 5ithin an industry setting, the presence of bentonite is beneficial for cuttings'carrying'capacity and filter ca/e characteristics. heological properties were investigated using a rotating viscometer %which is a type of diagnostic test for mud properties and thus performance&. 1t was determined that the mud samples were non'6ewtonian in character. Meaning that the apparent viscosity for the mud samples did not exhibit a direct proportionality between shear stress and shear rate.
"( The University of Trinidad and Tobago, Point Lisas Campus Esperanza Road, Brechin Caste, Couva!
") The University of Trinidad and Tobago, Point Lisas Campus Esperanza Road, Brechin Caste, Couva!
