AN EVALUATION OF SOME TILE.

DRAIN DEPTH
AND SPACING FORMULAE FROM THE PHYSICAL
PROPERTIES OF SOME ONTARIO SOILS1
F. R. HonB2 ,qNn D. A'{. Gn.tv3

On,tor io A gri cultural, Colle ge, Gu.el'ph, Ontario

[Receiled for publication February 13, 1957]

ABSTRACT
Soil ph1'5i621 properties cletennined on six hne-textured soil types near
Downloaded from www.nrcresearchpress.com by 137.59.223.73 on 01/29/20. For personal use only.

Bramotori and Chatharn s-ere used to soh'e and evatluate formulae for the
depth :rnd spacing
- of tile drains proposed b1' Hooghoudt, \\'alker, Slater,
and f)umu. At ihe Chatharn site, rneasurements oT weter-table drawdown
betu'cen tile drains rvere recorded to cornpzrre rvith Neal's water-table obser-
r-ations of adequate drainage.
A rvide difference in soil properties was exhibited between the two study
sites. None of the formulae results of spacing agreed precisely. Causes
for discreoancies betrveen the results rvere shorvn to be due to technical
clifferencei in appif ing datzr to the formula.e and to inherent differences in
derivatioir of the formulae. Data are presetrted shor.ing the need for a
phl sical criterion of adequ:rtc drainage.

INTRODUCTION

A problen-r of prime importance in tiie drainage design is the selection

Can. J. Soil. Sci. 1957.37:120-127.

oi the proper depth and spacing of lateral tile drains. A number ol studies
have beer-r made on this phzrse o{ design using various soil physical pro-
perties as criteria. In recent years hydraulic condttctivitl' and porosity
have been recognized as soil properties that directly goverll the placement
of drains. Drainage engineers, cognizart of the importance of these tu'o
properties, har.e made studies to derive methods for their determination.
This paper presents the results of two studies made to determine some
ph)tsical properties of six fine-textured soils in Ontario and an evaluation
of four depth ancl spacing- formulae utilizing these properties.

REVIEW OF T.ITERATURE
Because of the basic relationship set forth br, Darcy's Lar'r- betri-eetr
h1,-drauiic conductivit-rr and the velocitl. of r'r'ater lnovement torvards
drains, much study has been made to develop a suitable method to meilsttre
soil h1'd12slic conductivitl.. Uhland and O'Neil (12) have described a
direct method 'll'hereb]. r.rndisturbed soii cores are removed from the soil
and subjected to laboratory h1'draulic conductivity tests. 'lhis method
measures the conduc'tivit-v in a vertical plane. Luthin and Kirkham (5)
have reported on a piezometer method to measure the hydrauiic cottduc-
tivity in situ belo\v a s'ater-table. This method accents the horizontal
conductivity. Both methods are adzrptable for measuring the conductivitl-
of individual strata in a non-uniform soil.
- -lCo-rtribution from the Department of Agricultural Engincering, Ontario Agricultural College, Guelph,
Ont.
Assistant Professor of Agricultural Enginecring, Otrtario Agricultural College, Guelph, Ont
2

0 Formerly Graduate Assistant, Department of Agricultural Engineering, Ontario Agricultural College,

Cuelph, Ont.

t20
 -\ugust, 1957] HoRE AND GRAI-PHYSICAL PRor'MtrrES oF oN'fARro soILS 12I

Early attempts to derive depth al]d spacing formulzre \vere Inade by

zunker (15) and Neal (8) b). utilizing soil properties that indirectly affect
the hyclraulic conductivity. Zunker used the soil specilic gravit,v as the
criterion to determine the spacing of drains. 'fhis formula gives a limited
range in spacing because most mineral subsoils, u'hether rvell drailed
sands or poorly drained clal's, f ill have a specific gravity ranging betl'veen
2.6 and 2.7 (6). This narrou' variatiol iD specific gravity depends not
upon factors that affect the drzrinabilit-'- of a soil, but rather upon the
percentage of individual minerals of varying specilic grelvitlt present r'n
ihe soil. Neal related in nomograph form the depth and spacing of drains
Downloaded from www.nrcresearchpress.com by 137.59.223.73 on 01/29/20. For personal use only.

to moisture equivalent, clay content, a1d soil plasticity. It is qr-restionable

how far afield the nomographs u'ould appiy in view of other experiments
(2,7,l0) that have shown striking examples o{ the failure of these soil
properties to correlate reliablv r'vith hydraulic conductivitl'' Ne:rl recog-
th" limitations of usilg these l2boratory soil tests but stated that
"ir"d
they give a more accurate meals of detern'rining the proper depth and
spacii-rg than does a mere guess.
\,Iore recentllr, formulae in which h]'drauiic conductivity and porosity
enter explicitly 6a.'" been proposed b1'Hooghouclt (4), Slater (11), Waiker
(14) and Dumm (3). 'lhese attthors recognizcd the importauce of con-
sidering in their lormulae 1ot ortl-v h1'dratrlic colductivit-v but the amount
of urater that must be removed frorn the soil cluring il drau'do'tl'n period.
Can. J. Soil. Sci. 1957.37:120-127.

Basicailf it must be l<noryl hou' mrtch tvilter hils to be rcmoved to effect

a rvater-table recession thlt provicles :rdequ:rte dr:rinage. To clate, Neal's
observation of r':rter-tab1e conditions that pror.ided adequate draiuagc is
the onl-r- one reported in -\rnerica. He {otrr.rd that crops I'trere not seriously
injured if the n'ater-tnble u,as helcl 3t least 6 irrches belou, the ground
surface an<l nas lot'ered at the rirte of 1Ioot per day through the second
6-inch clepth-interval and at thc rate o{ 0.7 ioot per da-v through the third
6-inch clepth-interv:rl. '|1-ris observation, horvever, has not been verified
b)' organized research.
Fundameirtal anall'sis ci1' groulcl-g-ater movement ilto tile drains
shorn's that the flou' is essentialht horizontal rvhen the soil is underiain b1t
a shallorv impermeable la\rer, zruc] the flou' is radial ll'hen this la-ver is at
great depths. A combin:rtion clf the tu'o t1'pes of florv occurs u'hen the.
irnperme:rble la.ver is at an intermecliate depth.

MATERIALS AND PROCEDURE

'fhe first studl' 11ri15 made in 1953 at the Bramptol-r Seed Farm on
soils mapped as Peel cla-v loam (imper;t'ectly dra'ined'), Peel clay loam
(rnoclerately poorly dra'ined'), C:ristor clay loam ('imperJectly dra'ined'), Jeddo
cla1. fqarn (mod,erately poorly drained') and Jeddo clay loam (poorly d'rai,ned,)".
The Peel clay l6a- is a Grel--Brorn'n Podzolic soil developed in lacustrine
material underlain by cla-v ti1l at depths of 3 feet. The Caistor clay loam
is also a Grey-$16r.r'n Podzolic soil developed from clay till' The Jeddo
clay loam is a Dark Gre-v Gleisolic soil developed from clay till' Soil
physical measurements rvere rnade at the 1-, 2-, and 3-foot depths on each
+ Matthews, 8,C., and N. R. Richards, Soils report on Brampton seed farm, Brampton, Ont. Ont
4gr. Collcgc. 1919. Ulxbublished rerort.
 C\IIADIAN TOURNAL OF SOIL SCIENCE [Vol. 37

1l

€=> L
-.!
e*€
xg9< a:a c<{N
aat <reo\ o,oo <.o\
=

li o >:.= i
??? :?1 \a? N$N
iH-!vr

ll o I
J]
ri
l.
ti al
ll -.:'a^
6- / .' oN 14Nd
:il r:o='t=
d-+
r c <1 d +\oo\ NNi
N--!HNiOOc)c)c)

il .uYFtr
> ai5
oNo io I or40
^'^-^^.^^^^^^-'
000 0oo tar
Downloaded from www.nrcresearchpress.com by 137.59.223.73 on 01/29/20. For personal use only.

i *6 l
roj
i

rl

c
th
vG NA.t NON GlrN +rC OOO\ orl
\otl{9 o<'=! 44+ ro<r6 D4<tl

i
F

c '.:
=-::
>' I

7,
6 0F i q:- "l:9 19q a1a a"l:
o\
qY
iNi HiO iii trnN +rO cC\O
Can. J. Soil. Sci. 1957.37:120-127.

tr

li '6
3 4, .t-H i\CF
a
=a
-hN
9':? :\\
brN
::\ nt2n'. 1A:
NC N c
e
a+
mo ieH ddH :ii
=
h d
O
L
U
Fl
I

-^ *NO iNO iNC FNct iN.a -l

a
l

e.

F.]

t< 1l
o

l.;'c
a
!gL

o
x'r-13 a3

iO 'tr
>| '= L
o
=oi
a +
vsEEe
ddsEL E
1,

?AA a
qv-e
c.-!=c
tr=CC {
.,^
7.cc E
c
F-LOU
-:Ei:c
-{== a
v
 August, 1957] rroRE AND GRAy-pHysrcAr. r'RoprtRTrrls olj oN'rARIo so[s 123

soil type. At each location, under-the-water-table measurements of the

horizontal hydraulic conductivity were made in quadruplicate by the
piezometer method. The four piezometers were arranged approximately
6 feet apart in a square. Duplicate soil cores, 3 inches in diameter and
length, were obtained at the 1-and 2- toot depths, while single soil cores
only were obtained at the 3-foot depths. Difficulties encountered at the
3-foot depth due to presence of till virtually made the procuring of duplicate
samples an impossibility. I-aboratory studies on the soil cores rvere rnade
to determine the vertical hydraulic conductivity, bulk density, and per-
Downloaded from www.nrcresearchpress.com by 137.59.223.73 on 01/29/20. For personal use only.

centage of non-capillary porosity at 60-cm. tension using a techniqr-re

similar to that described by Uhland and O'Neal. Random soil sarnples
at each location were taken and the percentage of clay determined by the
Rouyoucos (1) hydrometer method.
The second study lvas made in 1955 near Chatham on a Brookston
cla1, i626 soll (poorly dra'i'ned'). 'Ihis soil type is a f)ark Grey Gleisolic:
soil developed from clay till (9). 'lhe same soil properties that were
investigated at Brampton were measured at the Chatham site, using
similar techniques except that the laboratory determinations on soil cores
and determinations of ciay content 'ivere rnade onl1' at the 1- and 2-foot.
depths.
Tlr'o tile drains, 50 feet apart and 2 feet deep, rvere present at the
Can. J. Soil. Sci. 1957.37:120-127.

Chatham site. These drains were installed in 1930 and v"'ere considereci
by the farm-owner to provide adequate drainage. Three-eighth inch-
diameter observation lvells rvere installed in a rolv at right angles to, and
at varying intervals fronr, the drains. The wells were spaced closer
together near the drains. During the drarvdow-n periods, the rvater-table
elevation between drains was measured with an electrical u,ater-level
indicator to determine the rate of drarvdolvn that provided, arbitrariiy,
adequate drainage.
Appropriate physical measurements rvere applied to solve formuiae
proposed by Hooghoudt, Slater, Walker, and Dumrn. Each formula
u.as solved by foilowing the procedure outlined by its author. Where
assumed values were requirerl, those suggested b)r the respective author
were used. Since the piezometer method ha-s the definite advantage of
measuring hydraulic conductivity ,in s,itu, these values v/ere used in the
Hooghoudt, Slater, and Dumfir formulae. Laboratory values as deter-
mined from soil cores \vere specifically required for the Walker formula.
Comparative drain depths of 2 feet and 3 feet were selected for the
formuiae calculations of spacing. The 3-foot depth u'as selected arbitrarily.
'Ihe 2-foot depth, however, r,vas chosen specifically to conforrn to the depth
of the installed drains at the Chatham site so that a comparison could be
macle betu'een the installed a.rrd calculated spacing in the Brookston soil,

RESULTS
'fable 1 lists the mean values of the physicai characteristics of ttre
six soil types. There is a considerabie difference in physical characteristics
between the soils at the Brampton and Chatham sites. Although the
percentage of cla-v is generally the sarne at the trvo sites, the soil at the
 124 C-A.\ADIAN ]OUR\^\I, OIt SOII, SCIE\CE [\ol i/

i.-
icr ^a\o N9
N-
c€ \cN
€4
61
N9
4
-r-
1G
E
aa al
a\
= '-

t5
= ctT -]C N€ -le N€
l!
i:
i-
Downloaded from www.nrcresearchpress.com by 137.59.223.73 on 01/29/20. For personal use only.

^l
.!
J (>o
r-N a dc) .ar
a ,! ctL
a I
:
:
- N€ NC NT' NO NM
L, 5
(.: I

U
n!
.f
Can. J. Soil. Sci. 1957.37:120-127.

7 N6

=
? !
a
5

c =d

a A .rc No N.o Nc 6ra E No

7 F,p
_a
c Ltl
3
4
z

O€ 4€
o+ iN iN NS

E
o
ir
I N€ NC N€ NS NO
N
-l

btr

!
'6

z
 ,\Lrgust, 1957] HoRE AND GRAY 't']HYSIcAL lRol'nR'rlti:s olr oNl-\RIo sort-s 125

Chatham site exhibits features, such ers lo'n'er bulk densitl' and higher
non-capillarS' porosity and hydraulic conductivitl', which indicate superior
soil structure.
Table 2 shon's the depth and spacing resutrts calculzrted {rom the four
formulae. All four lormulzre sholr. that the drain spacing increases lvith
increased depth in these soils. A comparison of the spacing results shows
th:rt none of tlie formulae results agrecrs u,ith another cousistently in all
six soil types. Generalll-there is better agreement between the Hooghoudt,
Slater, and Dumm results than there is betn'een these three formulae
results zrncl the Wall<er results. This could be expected because the
Downloaded from www.nrcresearchpress.com by 137.59.223.73 on 01/29/20. For personal use only.

hydraulic conductivit-v v:rlues r-tsed in the Wall<er formula .n'erc not the
salne as those used ir-r the other forrnulae.

DISCUSSION
An anal-u-sis of the discrepancies betu,een tlie resr-rlts for thc different
formulae shorvs that the causes of variance rvere due to diflerent techniques
of appl-ving soil physical dzrta to the formulae, and to basic differences in
the derivation of the forrnulae. Each autiror prescribed different methods
of applying the hydraulic conductivitl' data to the {ormulae. For example,
Walker suggested the use of the rninimum hl'draulic condr-rctivit;- value
to the drain depth; Slater suggested that the hl.clrauiic conductivitl' value
Can. J. Soil. Sci. 1957.37:120-127.

of the soil la1'er surrouncling the clrain shoulcl be used; ald Dun'rrn suggested
the use of the rveighted averaSlc h1'clraulic conductivitl- of the soil layers
above the barrier la-ver. 'I'hcre rt'ere :rlso different methods r-rsed to arrive
at a drainage rate to be :rppliecl to thc formulae. Only a suggested direct
drainage rate r':rs required for the IJooghoudt and Slater formulae whereas
a rvater-table recession rate itr conjunctiotr l'itl-r the soil porosity l1ras
required for the \lialker ancl Dumm formulae. The porosity factor logi-
cally is present in the latter trvo forlnulae to account for differences be-
trveen soils in the volume of rvater drained for a given urater-table recession.
Each zruthor, horvever, differed in his method of determining this volume.
Wall<er used the pore spzrce drained at 60-crn. tettsion, 'lthereas Dumm
used the specific yield.
Discrepancies in the results caused by basic differences in the derivation
of the formulae ma-v be attributed mainl-v to the t1'pe of florv assumed.
The Slater and Dumm formulae are based on the assumption of horizontal
flou' rvhile the Walker formula is based on radial florv. Slzrter, realizing
the limitations of his assumption when the barrier layer is at any appreciable
depth below the drains, sttggested that arn arbitrary barrier 2 leet belo'"v
the drains be assumed to adjust for convergcnce of flor'v near the drains.
Dumm did not consider that this refinement 'ivas perhaps ..varranted from
a practical viervpoinl. although convergcnce could be accounted for in his
formula. A slightly rvider spacing is derived 'w'hen local resistance due
to flou, convergence is neglected. Walker's assumption of radial flow is
valid only u'hen the barrier laysl- i. absent or at a great depth below the
drains. In addition to this limitation, the derivation of \A/alker's formula
implies that the hydraulic gradient causing florv toward a drain is unity-
This is not ahvays the case, for in practice it is usualllt less than unity.
 126 CANTIDIAN JOURNAL OF SOIL SCIENCE lVol. 37

Dtsfanee-Feef
20/oo/ozo30 50

y' sur,
,6ro< 'ce 1\
IE2B8
5.oof-

4
to/1tt

-1 SE 2i
tu \
zMat

4.@x
s
h
J// -,,
v \ =
Downloaded from www.nrcresearchpress.com by 137.59.223.73 on 01/29/20. For personal use only.

{, -rF s\
\ 3.OO
Frcunr 1. Drawdown cnn'es betrveen tile drains at the Chatham site for successive
days after a rainfall.

Hooghoudt's three formulae are based on the assumption of horizontal

flow, radial flow and a combination of horizontal and radial flow to meet
conditions rvhere the barrier layer is at drain depth, at a great depth below
the drain, and at some intermediate depth respectively. Hooghoudt's
formulae treat the problem of flow to tile drains more precisely than the
other formulae. Thus it is evident from the above discussion that there
are man)/ sources for variation in the results.
Can. J. Soil. Sci. 1957.37:120-127.

None of the formulae calculations of spacing shorrun in Table 2 at the

2-foot depth in the Brookston soil alarees with the actual installed spacing
of 50 feet. Agreement can only be obtained if the formula used for the
calculation is unique, ail the assumptions made in deriving the formula
are met, and vaiues of the variables entering into the formula are accurate
measurements of field conditions. Precise agreement in this study is,
therefore, impossible because none of the four formulae is unique; they
are at best only close approximations. The basic assumption of homo-
geneolrs soil conditions common to all forrnulae is not met in the Brookston
soil because of the presence of structurai fissures, and extraneous openings
caused by rvorms and roots. 'Ihe accuracy r','ith 'lvhich the soil variable,
hydraulic conductivitl', can be determirred is questioned by the authors
and has been questioned in the past by other u'orkers (3, 13, 14). For
agreement it is also required that the drainage rate used in the calculations
must tally vith the actual situation in the field. Unfortunatelf it u'as
impossiblc to rnersrrrc the cirainagc rale at thc st-udy site. Therefore,
drainage rates or related lvater-tab1e recession rates and porosity values
used in the formulae \vere those strggested bv the authors. Since Dumm
has no specific recession rate to offer for humid conditior-rs, Walker's stig-
gested rate was used in the Dumm formula. Ttrrese rates, of course, rvere
not necessarill' the sarne as the actual rate in the field. For example,
\\ralker used a recession rate of 0.7 feet per day through the third 6-inch
increment of soil in his formula. The drar,vdo\,'n curves in Figrire 1 shorv
that the rvater-table recession rate in the fieid \rras a great deal less than
that used in his formula.
From an applied point of vieu,-, it is of the utmost importance that a
drainage rate be used that rvill provicJe adequate drainage. Neal's obser-
rration of water-table conditions that provide adequate drainage is the
 August,1957l HoRE AND GRArr--pHYSrcAL PRoPERTTES oF oNrARIo soILS 127

only published criterion that has been expressed in physical terms. Figure
1 shows that at the Chatham site, where the farm-owner considered he u'as
obtaining adequate drainage, the water-table recession rate at the mid-
point between drains is a great deal less than that observed by Neal.
Thus it seems quite evident that more specifrc knowledge of the minimum
recession rate of the water-table for adequate drainage is needed.
It is doubtful whether the determination of depth and spacing by
formulae will become an exact procedure in the near future, because of
the heterogeneity of soils, climatic differences, and changes in economic
Downloaded from www.nrcresearchpress.com by 137.59.223.73 on 01/29/20. For personal use only.

conditions. However, an ttnderstanding of the basic principles involved

and more detailed information on the physical characteristics of soils
can be helpful to drainage engineers in making logical and reliable recom-
mendations.

ACKNOWLEDGEMDNT
The authors rvish to thank H. D. Ayers, Department of Agricultural
Engineering, Ontario Agricultural College, Guelph, for helpful criticism
of the manuscript.
REFDRENCES
1. Bouyoucos, G. J. A recalibration of the hydrometer n-rethod for rnaking mechanical
Can. J. Soil. Sci. 1957.37:120-127.

analysis of soils. J. Amer. Soc. Agron. 43 :434-438. 1951.

2. Childs, E. C., and N. C. George. Interaction of rvater and porous materials, soil
geometry and soil-rvater equilibria. Discussions of the Faraday Soc. 3 : 78-85.
1948.
3. Dumm, L. D. Drain-spacing formula. Agr. Eng. 35 :726-730. 1954.
4. I-Iooghoudt, S. B. Bijdragen tot de kennis van eenige natuurkundige groothederr
1an den grancl,'de'infiliratie
7, Algcrreeue beschorewing van het probleern van de detail
ontwaterir"rg en door middel van parallel loopcnde dreins, grappels,
slooten en kanalen. Versl. I-ancibouu'k Ond. XLVI : 515-707. 19+0.
5. I.uthin, J. N., and D. Kirkharn. A piezometer methocl of measuring permeability
of scil in situ below a s'ater table. Soil Sci. 68 : 349-357. 1949.
6. Lyon, T. L., and H. O. Buckm:Ln. The natnre and properties of soils' 4th ed.
NIacNlillan Co., Ner.v York, t\. Y. 1947.
7. Muskat, M. The flow of homogenecus fluids through porous media' J. !V' Edr,vards &
Sons, Inc., Ann Arbour, h{ich. 1946.
8. Neal, J. H. Proper spacing and depth of tiie drains determined by the physical
pioperties of the soil. I{inn. Agr. Expt. Sta. 'fech. Bull. 101. 1934.
9. Richards, N. R.., A. G. Cald*.ell, and F. F. l{orrvick. Soil survey of Essex County.
Ont. Scil Su;:ve-,- Rept. 11. Guelph, Ont. 1.919.
10. Roe, H. B., and J. K. Park. A study of the centrifuge moisture equivalent as an
indcx of the hydraulic permeabilitl' of satureted soils. Agr. Eng. 25 :381-385.
194+.
11. Slater, C. S. The depth and spacing of tile drains. Agr. Eng. 31 :448-450. 1950'
12. Uhiand, R. Il., and A. XL O'Neil. Soil permeability determinations for use in soil
and rvater conservation. U.S. Dept. Agr., S.C.S. Tech. Paper 101. 1951.
13. Visser, W. C. Tile drainage in the Netherlands. l'ietherlands J. Agr. Sci. 2 : 69-87.
1954.
P. Depth and spacing for drain laterals as computed from corc-sample
14. \\ralker, J.
perrneability lrleasuremcnts. Agr. Eng. 33 :71-73. 1952.
15. Zunker, F. Die Bestimmung der spezifischen oberfluche des Bodens. Landiv.
lahrb, l.VIiI : 159-203. 1923.