1975_12_1

Canadian Geotechnical Journal

Revue canadienne de geotechnique

Published by Publike par THE NATIONAL RESEARCH COUNCIL OF CANADA LE CONSEIL NATIONAL DE RECHERCHES DU CANADA

La rkponse hydrodynamique-ou des essais par injection instantanke dans un puit durant son dkveloppement-fournit une mkthode facile de mesure objective du dkveloppement du puit et de comparaison des diE6rentes methodes de dkveloppement qui peuvent btre utiliskes. L'kquipement d'essai par injection instantanke dkveloppk en Saskatchewan permet un en- registrement continu prkcis des changements de niveau d'eau en tous temps B I'exception des premieres secondes suivant I'injection dans le puit. Trois exemples d'analyse d'essais par injection instantanke indiquent que le jetting et le dkveloppement B l'air sont les mkthodes de dkveloppement les plus efficaces dans un sable aquifke du sud est de la Saskatchewan; au contraire, la mise en charge mkcanique renverse le processus de dkveloppement dans I'aquifer. La substitution d'une boue bentonitique pour de I'eau claire pour le forage de la zone finale dans un ~ u i t s'est avbrke nondksirable ~ u i s a u e des efforts et un temm beaucou~ ~ l u s considkrable ont

Volume 12 Number 1 February 1975 Volume 12 numCro 1 fivrier 1975

Hydrodynamic Response - or Slug Tests as a Means to Monitor the Progress of Well Development

J. A. VONHOF Water Resources Branch, Inland Waters Directorate, Environment Canada, 3303 33 St . N . W. , Calgaty, Alberta

Received November 1, 1973 Accepted August 1 , 1974

Sequential hydrodynamic r e s p o n s e o r slug tests in a well during well development-provide an easy method for objective measurement of well development and comparison of the various development techniques that may be employed. Slug-test equipment developed in Saskatchewan permits precise continuous recording of changing water levels for all times except the first few seconds after the introduction of the slug into the well. Three examples of slug-test analysis indicate jetting and air development to be the most successful techniques for well development in a sand aquifer in southeastern Saskatchewan; mechanical surging, on the other hand, actually reversed the development process in the aquifer. The substitution of a bentonite mud for clear water in drilling out the completion zone in one well proved to be highly undesirable because considerably more effort and time were required to develop this well.

ktinkcessaires pour dkvelopper ce puit.

Introduction During well construction some formation

damage inevitably occurs in the aquifer ad- jacent to the well bore. This reduces the per- meability of the aquifer adjacent to the screened interval or inlet section. This increases the drawdown in the well, the pressure drop at the formation/well interface and the entrance velocity of water into the screen. All of these factors increase the cost of well operation and

. * [Traduit par la Revue]

reduce the working life of the well. The objec- tive of well development is to restore the permeability of the formation contiguous to the inlet section to at least its original and preferably to a substantially higher value.

Because a slug test or hydrodynamic re- sponse test (Hvorslev 195 1 ; Ferris and Knowles 1954; Ferris et al. 1962; and Cooper et al. 1967) measures the transmissivity of an aquifer in the immediate vicinity of the well bore, such

Can. Geotech. J . , 12, l(1975)

2 CAN. GEOTECH. J. VOL. 12. 1975

tests run sequentially during well development providc an objective measure of the progress of well development. This report includes case histories of the results of sequential slug tests carried out during the construction of three ob- servation wells in southeastern Saskatchewan.

Theory The objective of well development is to bring

a well to its maximum designed production capacity. This is achieved by reducing the skin cffect, increasing the permeability, and creating a natural gravel pack in the immediate vicinity of the screened interval. The procedures and techniques involved in the development of screened wells in unconsolidated sediments are therefore designed to (Anonymous 1966, p. 294) :

1. Correct any damage to or clogging of the water-bearing formation occurring as a side effect from drilling (skin effect);

2. Increase the porosity and permeability of the natural formation in the vicinity of the well;

3. Stabilize the formation around a screened well so that the well will yield water free of sand, silt, etc. (creation of a natural gravcl pack).

The ultimate goal of well development is therefore to increase the permeability in the vicinity of the screened interval and to obtain a sand-free well.

The hydrodynamic response of a finite-diam- eter well to an instantaneous charge of water is a measure of transmissivity of the aquifer (Hvorslev 1951; Ferris and Knowles 1954; Ferris et nl. 1962; and Cooper et al. 1967). However, as Ferris et al. ( 1962, pp. 104-1 05) properly warned: "the duration of a 'slug' test is very short, hence the estimated transmissi- bility determined from the test will be repre- sentative only of the water-bearing material close to the well. Serious errors will be intro- duced unless the . . . well is fully developed and completely penetrates the aquifer".

Few wells completely penetrate an aquifer. However, according to Cooper et al. (1967) the vertical permeability of most stratified aqui- fers is only a small fraction of their horizontal permeability. If so, when flow is induced within a cylindrical volume of indeterminate radius around the well bore by adding an instan- taneous charge of water, then that flow will be

essentially radial. The aquifer thickness can be considered equal to the length of the inlet sec- tion and the transmissivity value can be con- sidered to represent that part of the aquifer in which the well is completed.

Because the volume of water added or re- moved during a slug test is small, the radius of investigation will be relatively small. For this reason a slug test is uniquely suited to measure the transmissivity immediately surrounding the inlet section, that is, in the region most affected by formation damage and well development.

The drawdown observed during a pumping tcst is the resultant not only of the transmissivity of the disturbed zone adjacent to the well bore, but also of the aquifer transmissivity near the well. The rate of drawdown in a pumping well after the effect of the volume of storage within the well becomes negligible is a function of the series transmissivity of concentric cylindrical elements of increasing radii from the well bore. Because a pumping test integrates the trans- missivity of a much larger cylindrical volumc around the well bore it is less affected by the effect of formation damage or well develop- ment.

In a properly developed well the transmis- sivity obtained from a slug test should be equal to or greater than the transmissivity computed from pumping test data. If there is still forma- tion damage the slug-test transmissivity would tend to be less than the aquifer transmissivity.

Equipment and Test Procedure The technique used for these response tests

was developed by Dr. W. A. Meneley, Sas- katchewan Research Council, specifically to determine the response characteristics of ob- servation wells (Meneley 1970). The equip- ment used was designed and constructed, or modified from commercially available equip- ment by R. Heinze, Saskatchewan Research Council. In the present study the instantaneous charge of water is produced by dropping a solid slug of known displacement into the well; the hydrostatic head decline is recorded with a conventional float-type water level recorder. Ferris and Knowles (1954) dumped a known volume of water into a well to approximate the instantaneous charge of water required by their mathematical model. They measured the head decline with a steel tape and "popper". Cooper qt al. ( 1967) produced an instantaneous head

VONHOF: HYDRODYNAMIC RESPONSE

change by instantly removing a float of known displacement from a well; they measured the head change by means of a pressure trans- ducer/recorder unit.

A solid slug of known displacement (Fig. 1B) is lowered to a position just above the water surface in the well on a steel cable. A modified Leupold-Stevens Type F recorder (Fig. 1A) is mounted over the well and the float is lowered until it rests freely on the slug. When the test is started the slug is released and allowed to fall at least 34 ft (1.1 m). As the slug falls the water level in the well rises rapidly lifting the falling float off the top of the slug. The recorder is started at the instant the slug is released and the declining head is traced as a curved line on the hydrograph chart.

The Leupold-Stevens recorder was modified by replacing the original clock and pen-drive mechanism with a set of gears driven by a 12V DC timing motor. One of four gear ratios can be selected to provide full travel of the pen across the chart in 4, 8, 16, and 32 minutes, respectively. A three-position FOR- WARD-OFF-REVERSE switch allows the di- rection of pen travel to be reversed if necessary to obtain longer duration rccords without sacri- ficing precision in time measurements (Figs. 3 and 4).

The slug (Fig. 1B) consists of a cylindrical aluminum body about 5 in. (-127 mm) in diameter, having an overall length of about 19 in. (-483 mm). It is fitted with three roller guides at the top and bottom of the cylinder. One guided wheel at each end is adjustable to allow the slug to be used in casing ranging from 6.25 to 7.5 in. ID (158.7 to 190.5 mm). The cylinder is weighted with sand to ensure rapid submergence. The guide wheels allow the slug to drop freely in the well when it is suspended by a cable fastened to the edge of the cylinder. This arrangement forces the sus- pension cable to the side of the casing where it does not obstruct the free movement of the float.

The recorder was mounted on a platform which could be easily mounted on the drilling rig (Fig. 1C) to facilitate testing during the development operations.

The equipment is applicable to aquifers having a transmissivity ranging from 0.1-lo2 cm2/s. If the transmissivity is lower than

FIG. 1. Slug-test equipment. (A) Battery-operated modified Leupold-Stevens Type F recorder. (B) Al- uminum slug. (C) Recorder installed on drilling rig.

4 CAN. GEOTECH. J. VOL. 12, 1975

0.1 cm2/s there is no advantage to using the solid slug because the error caused by non- instantaneous addition of a known volume of water is acceptably small. If the transmissivity is greater than 10Qm2/s the mechanical float- recorder system cannot follow or record the rapid decline of the fluid level. This equipment is limited. to use in wells having an inside diam- eter of at least 6.25 in. (158.7 mm) that are con- structcd sufficiently straight and plumb that the float drag on the casing is not excessive. The equipment is simple and sturdy and will operate reliably over the wide range of temperature (-50 O F to 110 OF) (-45.6 to $43.3 "C) en- countered in Saskatchewan. The equipment is rclatively inexpensive and for the most part it utilizes materials that are locally available. The instrumentation is stable and consistently pro- duces reproducible results.

Well Construction and Development This study was carried out during the con-

struction of observation wells in an unconsoli- dated glacial sand and gravel aquifer in south- eastern Saskatchewan. Test drilling and well construction were carried out by Elk Point Drilling Ltd., North Battleford, Saskatchewan, using a modified Failing 1500 hydraulic rotary rig.

The procedurcs used for test drilling and well construction were developed over many years by the contractor for successful well construc- tion in glacial aquifers (Topilka 1967).

Prior to well construction a test hole is drilled to determine thc texture, sorting, and stratifica- tion of the material in the completion zone. The selection of the completion interval is based on the description of the cutting samples, the single point resistance and spontaneous poten- tial electrical logs, and the driller's log. After this information has been obtained the test hole is abandoned and, if necessary, is plugged with bentonite-cement grout. The well is then con- structed within a 10-ft (3.05 m) radius of the test hole. This procedure is designed to elimin- ate formation damage resulting from drilling fluid entering the completion zone (skin effect).

The hole for the well is drilled to the top of the completion zone, casing is run, and the annulus is filled with bentonite-cement grout. After the grout has set (about 24 hours), the completion zone is drilled out using clean water as drilling fluid. The water is run off to waste

rather than being recirculated. A very detailed driller's log is kept of the drilling character, the color of the returning fluid (a measure of the silt and clay content) is noted, and samples are collected from each 1-ft (0.3 m) interval in the completion zone. After the completion zone has been drilled the screen assembly is made up using screens having the desired slot size and the necessary fittings and blank ex- tension pipe as the situation requires. Prior to casing installation a landing ring is welded to the bottom of the casing. This landing ring engages on the bushing at the box of the lead packer and prevents the screen assembly from passing entirely through the bottom of the casing. Following screen installation the posi- tion of the screen is verified by measurement with a weighted steel tape and then the packer is swedged. Well construction details are shown in Figs. 2, 5, and 6.

The development methods for the three wells had to be modified because the nonpumping water level was at or above the ground surface. The techniques employed include : high-velocity hydraulic jetting in the screen, air-lift pumping, mechanical surging, and hydraulic jetting with a dispersing agent (SAPP, sodium acid poly- phosphate). At the outset development pro- cedures are carried out gently to allow the sand to collapse around the screen. As develop- ment proceeds, the energy applied to the com- pletion zone is increased. Development is continued until the discharge water is essentially sand-free.

Case Histories Introduction

The results of successive slug tests carried out during the development of three observa- tion wells are used to illustrate the effect of development techniques and the rate of well development. Two wells show, after each suc- cessive development procedure, continuous in- crease in the transmissivity, whereas the third well shows a reversal of this trend. The first slug test on all wells was run immediately after screen installation and prior to any develop- ment. All wells show a drastic increase in the transmissivity after the first development pro- cedure.

To determine the transmissivity from the hydrographs (Figs. 3, 4), values of H / H o at various times are computed, where Ho is equal

VONHOF: HYDRODYNAMIC RESPONSE

LITHOLOGY

W E L L C O M P L E T I O N RECORD I i

Sand.,, obore

S m d a5 above .lo-IS% < gr sond Sond,m c g r , m o x . $ I ,ralc. suban9 wbr. -ell sorted

Sand 0 , .,bore

Sand a, .bore

C O N S T R U C T I O N I '$

M A T C H I N G CURVES-SLUG TEST ANALYSIS I AFTER COOPER r_l 9 , 1 9 6 7 )

E A C H T H E 0 R E : I C A L M A T C H I N G CURVE IS C H A R A C -

TERIZED BY A SPECIFIC VALLIE OF q WHERE

( 2 WHERE r~ = RADIUS SCREEN icml IICI .>s f c ' RADIUS WELL i c m l A N D

r: S : COEFFICIENT O f STORAGE

OBSERVED WATER LEVELS A N D BEST F I T T I N G THEORETICAL CURVES

'."'+....+r.... G I AFTER SCREEN INSTALLAT1ON

-0-._o_ G 2 AFTER AIR DEVELOPMENT - G 3 AFTER JETTING

--% ---- x-- G 4 AFTER AIR DEVELOPMENT

C A L C U L A T E D TRANSMISSIVIT IES

A F T E R SCREEN INSTALLATION I 1ocm7/,

AFTER AIR DEVELOPMENT 5.68 cd/% AFTER J E T T I N G 6 8 2 cml/>

AFTER AIR DEVELOPMENT 8 5 2 cm'/%

I, time since slug aubmwgsd lrecondr)

FIG. 2. Observation Well G. Slug-test analysis; test-hole and observation-well data.

to the maximum water level change at time t = 0 and H is the residual water level change at time t since submersion of the slug. The computed values are plotted versus time on semilogarithmic paper and the analysis to deter- mine transmissivity T is carried out by follow- ing the curve-matching procedure described by Cooper et al. (1967, p. 266). This analysis first requires the preparation of type curves for curve matching using the data tabulated by the same authors in their Table 1.

of 50 ft (15.2 m) below ground level in a medium- to coarse-grained unconsolidated sand.' The static water level is at ground level (Fig. 2 ) . The well is completed with two 4-ft ( 1.2-m) lengths of 4-in. ( 10 1.6-mm) diameter Johnson telescopic stainless steel well screens. Screen slot sizes are 0.012 and 0.015 in. (0.31 and 0.38 mm) for the upper and lower screen respectively.

After screen installation the well was de- veloped with air until the discharge water was

Observation Well G 'The sand size classification is according to Went- Observation Well G is completed at a depth worth Grade scale (Pettijohn 1957, p. 18).

CAN. GEOTECH. J. VOL. 12, 1975

1 rater level changcl

I

.................... G I AFTER SCREEN I N S T A L L A T I O N -.-.-. G 2 AFTER AIR D E V E L O P M E N T

% AFTER J E T T I N G ------- G1 AFTER A I R D E V E L O P M E N T

I 2 3

t, time since slug submerged (minutes)

FIG. 3. Hydrodynamic response hydrographs, Observation Well G.

sand-free. Total time of this operation was approximately 30 min. The transmissivity of the completion zone according to slug test G3 (Fig. 2) was 5.58 cm3/s. Secondly, the well was jetted with clean water for approximately 30 min to again achieve a sand-free discharge condition and the transmissivity increased to 6.82 cm2/s (G3, Fig. 2) . Subsequently the well was once more developed for approxi-

mately 45 min to sand-free discharge and the transmissivity showed an increase to 8.52 cm2/s (G4, Fig. 2) . No further development was done on this well.

To obtain a rapid and rough estimate of the effect of a well development procedure in the field the original hydrograph charts can be compared visually. Figures 3 and 4 show the reduced, superimposed original traces of the

VONHOF: HYDRODYNAMIC RESPONSE

- Ho. maximur

-.-.-. FI AFTER AIR DEVELOPMENT

Fa AFTER JETTING

F, AFTER SURGING - - - - F5 AFTER SAPP TRE

I MENT AND AIR DEVELOPMENT

I

1, time since slug submerged (minuter)

FIG. 4. Hydrodynamic response hydrographs, Observation Well F.

slug tests, and the increase or decrease in the lated. All slug tests show strong oscillations at transmissivity can be readily detected. It should the start of the test. According to Meneley be pointed out that the portions of the curves (1970, p. 47) this is caused by the relative representing the first 5 seconds are extrapo- movement of the float on the water surface.

CAN. GEOTECH. J . VOL. 12, 1975

VONHOF: HYDRODYNAMIC RESPONSE 9

Obsetvation Well F Observation Well F is completed in a fine-

grained, in places unsorted, clayey sand at a depth of 210 to 225 ft (64 to 68.6 m) below ground level. This well, which is located ad- jacent to Observation Well G , was constructed at this depth to obtain information on the hydrostatic pressure in the bottom part of the aquifer. The static water level is at surface.

Although the test hole (Fig. 5 ) showed medium to coarse-grained sand in the zone selected for well completion (the interval from 210 to 220 ft (64 to 67 m) below ground level), the sand encountered in the observation well over this depth range was fine grained. The test hole results suggested that the hole drilled for the completion zone should be ex- tended to a total depth of 240 ft (73.2 m) below ground level, in order to determine whether coarser sand was present below 220 ft (67 m) . At approximately 222 ft (-67.7 m) the sand became much coarser, but it contained considerable clay and silt, and furthermore ap- peared unsorted; the silt and clay content in- creased progressively with depth. The screen assembly was therefore designed for the interval from 210 to 225 ft (64 to 68.6 m) below ground level.

It was hoped that the bottom screen section, which was completed in a fine- to very coarse- grained, unsorted, silty, clayey sand, could be developed sufficiently to increase the overall hydrodynamic response characteristics of the well to an acceptable level. The wcll is com- pleted with three 4-in. ( 101.6-mm) diameter Johnson telescopic stainless steel screens. The top and middle screen are each 5 ft (1.5 m) long, whereas the bottom screen is 4 ft (1.2 m) long. Screen slot sizes are 0.007 in. (0.178 mm) for the top and middle screens and 0.015 in. (0.381 mm) for the bottom screen.

After screen installation the well was air developed for 1 hour. A large volume of very fine-grained sand was pumped out of the well. The transmissivity after air development was 1.75 cm2/s (F2, Fig. 5) . Subsequently the screen section was jettcd with clean water for 2+ hours, the jetting being concentrated on the bottom larger-slot screen. Sand larger than the 0.007-in. (0.178-mm) slot size of the upper screcns was pumped out. This indicated that development of the lower screen was progres-

sing favorably. The transmissivity increased to 2.35 c m y s (F,, Fig. 5 ) . To further improve well development it was decided to surge the well mechanically. After each 15-min period the well was pumped by air-lift. The discharge was dirty and seemed to contain a high per- centage of clay. This procedure was repeated several times. A slug test at the end of this procedure, however, showed a drastic decrease in transmissivity (Fg , Fig. 5 ) to 1.58 c m v s from 2.35 cmvs . To correct this the screened intcrval was jetted with 300 gallons ( 1363.8 1) of clean watcr containing a clay dispersing agent (SAPP) and was left overnight. The next morning the well was air developed for 45 min. The transmissivity increased to 2.06 c m v s (F,, Fig. 5 ) . Further air development did not materially change the transmissivity. A slug test (not shown) run after a pump test at a rate of 50 gallons per minute (225 l/min) for 16 hours did not show any marked improvement in the transmissivity.

Observation Well D Observation Well D is completed at a depth

of 33 ft (10.1 m) in fine-grained sand. The static watcr level is 7 ft (2.1 m) above ground level. The well is completed with two 5-ft (1.5-m) lengths of 4-in. ( 101.6-mm) diameter Johnson telcscopic stainless steel wcll screens. The screen slot size for both screens is 0.007 in. (0.178 mm). The observation well was con- structed 8 ft (2.4 m) from the test hole. Con- siderable difficulty was experienced in develop- ing this well. On the basis of the electric log and the samples, medium- to coarse-grained sand was expected to be present in the com- pletion zone of the observation well. It was therefore decided to determine the effect of drilling out of the completion zone with a heavy bentonite mud rather than with clean water. However, during the drilling of the completion zone it was found that only fine-grained sand was present. After the screen was installed the well was bailed dry and the effect of the drilling mud became evident. The water level after half an hour waiting was still at 25 ft (7.6 m) be- low surface. The next three hours were spent on air development, jetting with watcr and SAPP treatment. After these procedures the static water level was approximately 2.5 ft (-0.8 m) above surface. A slug test (D,, Fig.

10 CAN. GEOTECH. J . VOL. 12. 1975

MATCHING CURVES-SLUG TEST ANALYSIS (nrrra coom 11 n_t.l~a7 I

g : 10-' FOR * i t MATCHING CURYPI

OBSERVED WATER LEVELS A N D BEST FITTING THEORETICAL CURVES

AFTER A l e D fY f tOPMfN I . ....-.....*.....+. 0 2 lE lT lNG i iND S A W IR fATMfN I

d.9 D l AF l f i l f t O W $ N G OVER NIGHT

eel. D l i i i l f l l I f I T I N G .---.--- 01 111111 l l R D f V f I O P M f N I

I . lime since duo wbrnmrspd I seconds 1

FIG. 6. Observation Well D. Slug-test analysis; test-hole and observation-well data.

6) indicated a transmissivity of 1.03 cm2/s. No further work was done that day and the well was left flowing at a rate of approximately 3 gallons per minute (14 l/min) overnight. The next morning another slug test (D3, Fig. 6) was run which showed an increase to 1.14 cm2/s. Subsequently the well was jetted for 3 hours until no traces of sand and clay could be found. Analysis of slug test D4 (Fig. 6 ) indicated the transmissivity had increased to 1.45 cmvs. Finally the well was air developed for another hour, which resulted in an increase in trans- missivity to 1.95 cm2/s (D5, Fig. 6 ) .

The extent of the damage caused by drilling the completion zone with a bentonite mud can- not be determined, because no second well was completed by the clean-water procedure in an identical fine-grained horizon. However, the difficulties encountered during well develop-

ment suggest an adverse effect of drilling mud during the drilling of the completion zone and should be avoided.

Figure 7 summarizes the well development procedures in the three observation wells. The effect and rate of progress of well development can readily be seen. The most suitable tech- niques for this type of aquifer appear to be jetting and air development.

Conclusions Hydrodynamic response evaluation by slug

testing during well development measures sys- tematic changes in the transmissivity of the zone adjacent to the completion interval. Slug tests provide a quick and easy method of mea- suring the overall effectiveness of well develop- ment. The measurements should be used in conjunction with additional bore-hole measure-

VONHOF: HYDRODYNAMIC RESPONSE 11

0.00

Observation well D

Observation well F

0 0 - ./

2.50

2.00

1.50

1.00

0.50

SCREEN INSTALLATION

SCREEN INSTALLATION AIR DEVELOPMENT JETTING AIR DEVELOPMENT

SCREEN AIR DEVELOPMENT INSTALLATION JETTING AND SAPP

TREATMENT

/ -\ & / / .

4.- ,xR

'. /---

'LX/--

/ /

-/

/ /

/

Observation well G

FIG. 7. Effect of well development procedures.

AIR DEVELOPMENT

ments, such as, for example, gamma logs to struction and development and the results from obtain objective data about the effectiveness of successive hydrodynamic response tests indicate any particular development technique. that jetting and air development are the most

The experience obtained during well con- successful development techniques; mechanical

FLOWING OVERNIGHT

JETTING

JETTING AIR DEVELOPMENT

SURGING SAPP TREATMENT 8

AIR DEVELOPMENT

12 CAN. GEOTECH. J. VOL. 12, 1975

surging in this aquifer, on the other hand, actually reversed the development process.

The use of bentonite mud in drilling out the completion zone is undesirable and should be avoided, because considerably more effort and time are required to develop the well. The introduction of clay dispersing agents in the completion zones of those wells where clay is present in the aquifer or where clay is intro- duced by drilling mud has beneficial results for well development.

Acknowledgments The writer wishes to express his sincere

appreciation to Dr. W. A. Meneley, Saskatch- ewan Research Council, for his valuable advice and critique in the preparation of this paper.

Thanks are due to both Dr. Meneley and Dr. J. A. Cherry, University of Waterloo, for critically reading the manuscript.

ANONYMOUS. 1966. Groundwater and wells. Edward E: Johnson, Inc., Saint Paul, Minnesota.

COOPER, H. H. JR., BREDEHOEFT, J. D., and PAPA- DOPULOS, I. S. 1967. Response of a finite-diameter well to an instantaneous charge of water. Water Res. Res., 3, pp. 263-269.

FERRIS, J. C., and KNOWLES, D. B. 1954. The slug test for estimating transmissibility: U.S. Geol. Surv., Ground Water Note 26.

FERRIS, J. C., KNOWLES, D. B., BROWN, R. H., and STALLMAN, R. W. 1962. Theory of aquifer tests. U.S. Geol. Surv., Water Supply Paper 1536-E, 174 p.

HVORSLEV, M. J. 1951. Time lag and soil-permeability in groundwater observations. Waterways Expt. Sta. Bull. No. 36, Corps. Eng., Vicksburg, Miss., 50 p.

MENELEY, W. A. 1970. Groundwater resources. I n Chris- tiansen, E. A. (Ed.), Physical environment of Saska- toon. Nat. Res. Counc. Can., Ottawa, pp. 39-48.

PETTIJOHN, F. J . 1957. Sedimentary rocks (2nd edn.). Harper and Brothers, New York, 718 p.

TOPILKA, J. D. 1967. Farm water wells. Water Stud. Inst., Saskatoon, Sask., Rep. No. 3, pp. 36-40.