AnExplanationofmyOllaIrrigationSystem

ByTomBowes,Ver.1.0Feb.2012

IvehadafewquestionscomeinviaemailregardingthepresentationthatI
did for the 2012 Ann Arbor Reskilling Festival on February 4
th
. Some of you
have managed to pick up on the course handout via the web and have
wanted some clarification about my project and its methods. Let me say
rightupfrontthatIvealreadychangedsomethingsbasedonthedesireto
simplify the system. This written explanation will essentially take you from
top to bottom, literally, and tell
you what I have learned thus far
about putting together a self
feeding, subsurface watering
system based on using olla (oh
yah) capsules as the means of
deliveringwatertogardenplants.

TheOverallSystem

Conceptually, what Im trying to

do with my olla irrigation system
is to implement, at a practical
level,oneofthecoreprinciplesof
Permaculture; to capture and
slow the flow of water. The system which Ive designed is capable of being
fed via rainwater in a gravity fed arrangement, or via a garden hose. Using
graywaterwouldalsobepossible.Thefundamentalsofthissystemarewell
documentedbyalonghistoryofvariouscultureswhichhaveusedollasor
unglazed clay pots, filled with water and buried next to plants in a garden.
Thepervious(yes,thatisaword)claypotsallowthewatertoseepthrough
theirwallsataveryslowrate,providingirrigationtotheadjacentplants.

Traditionally, the ollas would be filled and then covered with a piece of
stoneinordertokeepmosquitoesfromusingthemasabreedingplace.The
daily regimen would then be to refill each of the ollas, which I tried, but
decided was too much work. What I have done is to experiment with a
method of providing a continuous water feed to the buried olla capsules,
thereby reducing waterrelated stress on both myself and the plants in my
garden.

In reading my missive, keep in mind that this project is indeed an

experiment and is subject to change. I will attempt to update the
informationfromtimetotime,andreleaseitviatheweb;butiftheresone
thingthatIvelearnedinpublishingarticlesviathewebitsthatonceitsout
there,youcanttakeitback.Feelfreetogivemeyourfeedback.

TheRainBarrel

LastsummerIhadtwoollasystemsatmyresidence.Oneusedarainbarrel
to capture runoff from the roof of my house, the other used city water.
Since then Ive figured out how to build a hybrid system that will use both
sources, which Ill explain later. My
rain barrels were a carryover from
previousyearsandoneofmygoals
was to be able to use them as part
of the olla feed system. I was able
to implement this, but only after a
failed attempt using a toilet valve
as a regulator. Again, Ill explain
later.

The total cost for my pair of rain

barrels was about $50, and this is
only because I spent the big bucks
and bought two premade
bulkhead adapters to tie the two
barrels together, and to connect
the hose bib. If I were doing this
again Id save a few dollars and
make my own bulkhead fittings
from some PVC electrical
adapters. There are a couple of
videos on YouTube which describe
the process of making a bulkhead
adapter, or you can find them
onlineforabout$10each.

One of the nice things about using

ollas for irrigation is that they are
fairly impervious to having some
particulate in the water supply.
Whatthismeansasapracticalmatteristhatyoudontneedtorigupafirst
flushapparatustokeepthegunkoutoftherainbarrel.Infact,allofthebird
poopandotherstuffthatyouandIwouldntwanttodrinkworksfinewhen
watering plants. My rain barrel uses a simple piece of aluminum window
screen as a coarse filter to catch the major debris, such as the seed pods
frommyneighborsmapletreeasyoucanseeinthephoto.

By connecting the two trash cans together the capacity of the rain barrel
systemisdoubledto66gallons.Ifoundthatthe33gallontrashcansgoon
saleoccasionallyforabout$810,whichispretty cheapwhen compared to
someoftherainbarrelkitsonthemarketwhichcostupwardsof$100.You
couldalsousesurplus55gallondrumswhicharesometimesavailablefora
lowprice,butonehastochecktobesurethatthepreviouscontentsofthe
drumarenottoxicorundesirableforgrowingfood.

Anadvantageofusingthetrashcansisthattheycanbestackedandstored
during the winter, taking up less space than their industrial drum
counterparts.

One of the other things that my

rainbarrelsetuphasisanoverflow
fittingatthetopofthebarrel.This
issimplyafittingforagardenhose
and allows usage of the excess
water which sometimes comes
along in a good rain event. By
routing the excess water to larger
olla capsules located near plants
which have a lower water
requirement the benefits of
rainwater harvesting can be
extended beyond the traditional
garden.

Onecaveattothisisthatthebleedertubesfortheselargerollasshouldbe
left unplugged (only on ollas which will be fed sporadically) so that the
capsuleswill fillreadilyandnot bestopped byexcessivebackpressure.(Ill
explain this in a bit.) This excess water can also be routed to heavily
mulchedbedstobetterutilizeit.

TheRegulator

ThewaterfeedbucketorregulatoristhepartofthesystemthatIthinkI
can lay claim to. In all of the videos and articles that I watched and read
prior to setting up my system there was not one which used this method.
Theconceptisquitesimple;bymaintainingconstantwaterfeedpressureon
the olla capsules, the system will continue to run unattended. Basically,
onceyouhaveeverythingsetupandprimed,theregulatormakessurethat
water will be constantly delivered to all of the olla capsules, even if one
springs a leak. It also provides for a point to supply liquid fertilizer to the
system.Theregulatorbucketsitsonthegroundinthevicinityoftheollasto
beserved.
The constant water source and feed pressure is achieved by virtue of
something called head pressure. As an instrumentation technician Ive
had it drilled into me that for every 27.7 inches of depth in a container of
watertherewillbearesultantpressureatthebottomofthecontainerof1
PSI. The pressure is proportional to the depth of the water, which means
that with the average 5 gallon plastic bucket filled with about 10 inches of
water,thepressureatthebottomofthebucketisabout.38PSI.

Control of the water level in the bucket is achieved by means of an

evaporative cooler float valve. (Dial Mfg. #4164plumbersstock.com)
Connected to the valve is a dishwasher elbow (Watts A158A Lowes)
which allows a generic
garden hose to be
connected to supply
watertothebucket.

Initially I experimented
with a more readily
available toilet valve,
installed in the bottom
of the bucket, but
foundthatitwouldnot
operate with the low
pressure supplied by
the rain barrel. The
evaporative cooler
valveworkswitheither
the low head
conditions supplied via
a slightly elevated (12
16 above ground) rain
barrel or with the high
pressureofacitysupplyorwellpump.Apieceof3/8coppertubingisused
tocouplethecompressionfittingsofthevalveandthedishwasherelbow.

OnmyoriginalsystemsIconnectedalengthofI.D.by.7O.D.irrigation
tubing to the bottom of the bucket by means of inserting the tubing into a
rathertight11/16hole.Ivesincerethoughtthisinanefforttosimplifythe
installation and reduce the number of different parts needed. I found
through testing that a length of irrigation tubing can deliver about 10
gallonsperhourfromtheregulatorbucketunderaheadpressureof10of
water.Therefore,Imnowrecommendingthatyousimplydrillaholeinthe
sideofthebucket,about1abovethebottom,usingasizeC(.242)drill
bit. The irrigation tubing can be inserted directly into the hole without
additionalfittingsandwillholdfastunlessafairamountofforceisexerted
to pull it out. Use black polyethylene (not vinyl) irrigation tubing and it
shouldlastaminimumofsevenyears;probablylongerthanthebucket.

I had mentioned
earlier that Ive
figured out a way to
supply the system
from both a
rainwater and a
pressurized water
source. This can be
done by adding a
second valve on the
other side of the
bucket and
adjustingthefloatleveltoalowerheightforthepressurizedwatersource.
In essence, you would set the rainwater float to maintain a level of about
10 and the city or well water source at about 6 above the bottom of the
bucket.Thatwayiftherainbarrelrunsdry,thelowerfloatandsecondvalve
will regulate the city water at about 6 until the rain barrel gets filled back
up again. An empty cat litter bucket works great for this because there is
enough room for both valve floats. Id recommend using a backflow
preventer on the pressurized water line if you decide to use a dual feed
system.

Additional holes can be drilled for more feed lines, allowing a single
regulator bucket to be used for multiple water circuits. If you need to
abandononeofthefeedlinessimplycutitoffandplugtheendofthetube
withagoofplug.

TubingandStoppers

In my initial experiments with connecting PE irrigation tubing to my olla

capsules I attempted to simply caulk the tubes into the holes in the top of
the ollas using silicone caulk. The arrangement was unsatisfactory and
resulted in a lot of leaks, and tubes which slipped out of the holes. The
problem is that the surface of the tubing is quite slick and the caulk just
wont hold it. Ive seen on the Internet where another experimenter was
using epoxy adhesive to glue a couple of hose barb fittings into the hole in
the olla capsule. The thing which I dont like about that approach is that
long term exposure to uncured epoxies creates severe allergic reactions in
somepeopleandifthefittingsbreaktheassemblywoulddifficulttorepair.
Ichosetouserubberlabstoppersasanalternativeandtheyseemtowork
quitewell.Usingatwoholelabstopperasthesealingmeansatthetopend
of the olla allows me to change tubing and fitting configurations, manually
filltheollas(handyincaseofarepairoraddition),inspectthecapsules,and
introduce specialty fertilizers for specific plants. The stoppers are fairly
inexpensiveandshouldlastformanyyearsbeforeneedingreplacement.

The photo at the left shows the basic components to connect the
irrigation tubing to the ollas. Ive boiled it down to three types of fittings
and one type of stopper to simplify things. The entire system can be
plumbed using only tee and straight hose barb fittings and end plugs (also
calledgoofplugs).Allofthesefittingscanbefoundatahydroponicsupply
store or ordered online (irrigationdirect.com). The rubber stoppers,
however,willprobablyhavetobeorderedonlinesincemostcitiesseemto
be lacking a scientific supply store. Be sure to order the black rubber
stoppersandnotthenotthenaturalgumrubberstoppersasthey(theblack
ones)areUVresistantandwilllastmuchlonger.

Thereare(obviously)differentsizesoftwoholestoppers.Thesize0stopper
isthesmallestonewhichcanbepurchasedwithtwoholes.Iusesizes0,2,
4, and 6, which as shown give a tapered continuum when stacked end to
end. This allows me to connect to a pot with just about any size hole from
toabout11/4.Thehosebarbfittingsareinsertedintotheholesinthe
stoppers, with bits of the
tubingpushedontotheendof
the barb to provide a tight fit
forthelargerholesinsomeof
thestoppers.

Here a sleeved tee fitting and

astraightpieceoftubinghave
been inserted into a size 6
stopper. Part of the strategy for my system of interconnecting the olla
capsulesistohaveafeedandableedforeachone.Thefeedandbleed
arrangement allows the water to enter via one hole and the air to escape
viatheother,whichishelpfulwheninitiallyfilling,andlaterdiagnosingthe
system.Theindividualcapsulesmayshareafeedlineviateefittings,butthe
bleedlinesremainseparate.

Ivefoundthattherubberplugstendtosealupthebestwhentheycanbe
insertedintotheholeinthepotsothatabout2/3ofthestopperisdownin
thehole.Thisisespeciallyimportantwiththesize0stopperssincepushing
the bottom half of the stopper against the sides of the hole can actually
pinch off the water flow in some cases. What this means as a practical
matter is that youll
have to match the size
of the holes in the top
ofthepotsyoureusing
to the size of the
stopper. A size 0
stopper works best
witha9/16hole,size2
11/16, size 4 7/8,
andsize61and1/8.
Therearealsooddsizes
of stoppers available, i.e. 1357. You can find information about stopper
sizesatmcmaster.combysearchingforthetermrubberstoppers.

ToopenuptheholesinthetopendoftheollasIuseeitheracarbidedrill,
chucked up in a drill press, or an oldfashioned pipe reamer. The reamer
works best when flushed with a bit of water. Ive also used a battery
powered drill to turn the carbide bit, using great care to go very slowly
when enlarging the hole. I find that it works best to open up the holes to
theirfinalsizebeforeassemblingtheollacapsules.Theholesinthebottom
oftheollacanbeleftattheiroriginalsmallersize.

Ive purchased most

of my stoppers
online at
labandscience.com.
The stoppers are
sold by the pound,
and there are about
75size 0 twohole
stoppers to the
pound. My estimate
is that by the time
you get to size 6
there are only about 20 per pound. Ive found that even though all of the
stoppersaresupposedlymadeofthesamestuffandpresumablycomefrom
thesamesourcethereisawidevariationonthehardnessoftherubberand
the size of the holes. It doesnt seem to make a difference, other than the
harder stoppers need a bit more seating force, and the ones with larger
holes sometimes need a piece of tubing slipped over the hose barb before
insertion.

Ive found that when

connecting a string of
ollas together its best
toadjust thetubingso
that curvature of the
tubing is in the
horizontal plane, as
shown below. If the
curvature is in the
vertical plane it tends
to pop the stoppers
out of the ollas. Also,
its easier to assemble the tubing, hose barbs, and stoppers as much as
possiblebeforeinsertingthestoppersintotheollas.

TheOllaCapsules

Just about any size of unglazed clay pots can be used to make your olla
capsules.Thedeterminingfactorsastosizeboildownto:

Howmuchroomdoyouhave?
Whatsizepotsdoyouhave?
Whatcanyougetforfree?
Whatareyoutryingtogrow?
Doyouhaveaconstantwatersource?

In the photo above you can see a variety of olla capsules which Ive made
from varioussizes of pots. With the exception of the second largest set, all
of these pots were gathered for free from a local charitable organization
that runs a donationbased resale shop. When buying pots I tend to stick
with the ones which are about 4 inches in diameter at the rim since they
seemtobethebesttradeoffbetweensize,easeofuse,andaffordability.
YoumightwonderwhyIwouldchoosetouseeitherverylargeorverysmall
pots.Theansweristhateachofthemhastheirapplication.Thesmallones
can be used very effectively when container gardening or growing shallow
rootedplants.Thelargeonesarebestinsituationswherethewatersupply
is sporadic since they can be used to capture rainwater and then slowly
releaseitintothesoil.Thelargeonesarealsogreatforplantswhichrequire
alotofwater,suchasmelonsandsquash.Oneothermethodofutilizingthe
smallpotsistoburythembelowthesurfaceofthesoilinordertogetthem
downintotherootzoneoftheplants.Solongastheyhaveaconstantwater
feedtheywillbenearlyaseffectiveastheirlargercousins.

Infact,Ididanexperimentwiththesmallestoftheollasshown,whichhave
only a 2.5 inch diameter. I was looking to provide water to a raised bed
whichis4x4insizeanddecidedtotrynineofthesmallollas,spacedat16
in a 3x3 configuration. When connected to a regulator bucket with about
10ofheadpressureIfoundthatthe flowratetothenineollaswasabout
20mL per minute or about .3 gallons per hour (GPH). This is a pretty
considerable amount of water, given that it was a bench test, without the
osmotic action of the plant roots to help draw water through the walls of
the ollas. My guess is that the thin walls of the pot let through a greater
amount of water relative to the size of the pot, but I need to do some
further experimenting to determine the wall thickness to water delivery
characteristics.

Myplanwiththeseistouseasingleirrigationtubetofeed4raisedbeds,
withatotalof362.5ollacapsules.Thecapsuleswillbeconnectedtogether
inonestringforeachbed,withateefittinginthesupplylineforeachbed.
By having the bleed tubes stick up above the surface of the soil I can
determinethepositionofeacholla,andbyremovingtheplugfromthelast
bleedtubeinthecircuitIcandeterminetheintegrityofallnineoftheollas
ineachbed.BecausethesitehaswellwaterIcangetawaywiththesmaller
capsulesandutilizeaclosed,pressurizedcapsulesetup.

The ollas capsules are nothing special; simply two unglazed clay pots glued
together,usinghighqualityblack(UVresistant)siliconecaulkandapieceof
ceramicorglasstilegluedintoplugtheholeinthebottompot.

Thedrillbitshowninthephotowasusedtoremovetheburrfromthehole
insideofthepotsothatthetilewouldsitflat.Itcouldalsobeusedtoopen
uptheholeinthetoppottoprovideforthecorrectinsertiondepthforthe
rubber stopper. The ceramic tile was a mere $2.50 at the local builders
supply and provided enough tiles for about 50 ollas. (Ive since mostly
switchedtousingsheetsof1squaretiles.)The4potsrangefrom$.65to
$1.25locally.

Assemblyoftheollabeginswithgluingapieceoftileinto plug theholein

thebottompot,andthenapplyingabeadofcaulktotherimofthebottom
pot. The top pot is placed onto the bottom one and the excess caulk
smearedaroundtosealupthejoint.OnethingthatIhavefoundishelpfulis
to dry fit and align the rims of the pots prior to caulking. Putting an
alignment mark on the pots with a pencil or marker makes it easier to put
thepotstogetheroncethecaulkisapplied.Somepotsareseriouslyoutof
roundandyoucantmakethemlineup,regardlessofhowmuchyoutry.If
youarebuyingpots,inspectthemforcracksandroundnessatthestore.Its
not unusual for a fair percentage them to be defective, and its a hassle to
have to go back and exchange them later. (Cracked ollas will leak a
surprisingamountofwater.)

WhenIburymyollasItendtodigtheholeabout2deeperthantheollais
tallifIwantthetopoftheollatobeflushwiththesurface,and4deeperif
Iwantitbelowground.Iliketoputabitofcompostandasmallamountof
organic fertilizer into the fill dirt which will go back in around the olla. My
thought is that this creates an ideal environment for the roots of the plant
wherebothmoistureandnutrientsareavailable,andincreasescropyield.

QuestionsandStrategies

Afewpeoplehaveaskedwhetherornottheollascanbeleftintheground
inaclimatewherefreezingisafactor.AtthispointIcanonlysaythatIshut
offthewatersupplytomysystemlatelastOctober,lefteverythingwhereit
was, and Ill know in about two months how everything fared. Supposedly
thefreezingofthemoistureinthewallsofthepostwillcreatesomecracks,
butthisispartofmyexperimenttotryanddeterminejusthowmuchworkI
can get by without. At least with the rubber stoppers I can fill each olla
separatelyandseeifIhaveanyleaks.

Some of you community gardeners have asked how I would adapt this
system to your situation. My thoughts on this are that I would determine
whether or not you can establish a constant water supply source, using
barrels to store the water above ground. In that case the system could be
installed as Ive described using smaller ollas, in the 4 range. This would
enable multiple ollas to be connected together using a single run of
irrigation tubing to the regulator, thus minimizing the amount of irrigation
tubingneeded.

Lets say that you dont have a continuous water supply, but can leave a
bucketatthesitewherethefeedlinesoftheollascanremainconnected.In
thiscaseIwoulduselargerollas,inthe6to8rangeandrunanindividual
feedtubefromeachollabacktotheheaderbucket.Ifthereisnorainwater
harvestinginvolvedyoucouldgetbywithouttheregulatorvalveandsimply
fillthebucketwithhauledwatereachtimeyouvisitthegarden.Inthiscase
makethebleedtubeslongenoughtoreachuptoalevelwhichisjustbelow
the top of the bucket and leave them unplugged. When the water spouts
outofthebleedtubesyoullknowthesystemisfullandyoucanstopfilling
thebucket.
If you can arrange to have a larger secondary reservoir which is still
intermittent then were back to using the regulator valve, but sticking with
the individual supply lines to the ollas. The rationale here is that the larger
ollascanstoremorewater,essentiallycreatinganundergroundreservoir.If
you cant leave the bucket out in the open, youre back to filling the ollas
individually; pull out the rubber stopper and proceed, the old fashioned
way.

How do you prime the system? is a question that Ive been asked more
than once. In a nutshell, you connect the regulator bucket to the water
source, and starting with the olla closest to the bucket, either remove the
plug from the end of the bleeder tube or loosen the stopper in the top of
the olla. When the capsule is full, as evidenced by water coming out the
bleedertubeorfromthetopoftheolla,youcanreinserttheplugandgoon
to the next olla down the line. Keep working away from the water source,
removingtheairfromthesystemasyougo.Onceeverythingisprimedyou
can walk away and the system will maintain itself, provided you dont run
outofwater.

Ollas can reduce water consumption by up to 80% and increase yields

significantly. They can also give your garden the resilience lacking when
thingshavetobewateredbyhand,byyou.Thisisinadditiontosavingthe
laborofwatering.

AsWayneWeiseman,myPermacultureinstructornoted,Youhavetoputit
ontheground.Iveseenenoughofthissysteminactionnowtoknowthat
it makes a huge difference in the amount of labor for, and the health of, a
garden.Iwentfromcominghomeatnooneachdayandseeingmytomato
and squash plants all wilted in the noonday heat, to one week after
installing ollas having a robust and healthy garden with no heat stress on
theplants.

If you have comments or questions about my system, or any additional

insights, please feel free to ask via the Midwest Permaculture website
(http://midwestpermaculture.ning.com/group/designsandprojects/forum),
orviaemail.Myemailaddressis:

Bestwishes,

TomBowes