Post #G21-008: Building a plumbed core bed.

Posted on April 13, 2021


Today it took me about an hour to convert a 4′ x 16′ raised bed into a “core bed”.  That’s a bed with a considerable amount of straw buried in a trench running down the middle of the bed.  In theory, the straw core will retain water and reduce or eliminate the need for watering.  I’ve set it up so that I can pipe water directly into the core, so that if I do need to water it, I can do that in next-to-no-time.

If you just want to see the construction, go to the next section.


I first heard about core beds from the MI Gardener, a relentlessly cheerful presence on YouTube.  His gardening approach (“The Autopilot Garden”) focuses on minimizing the effort required to maintain a garden.  This self-watering “core bed” is an integral part of that.

Having an underground reservoir built into the bed appealed to me for reasons that are unique to my situation.  I summarized those in my just-prior garden post.  Briefly:

  • I want to irrigate some raised garden beds.
  • I want to use rain water from water barrels located some distance from those beds.
  • I have more-or-less zero water pressure from those barrels.
  • I don’t want to run a water pump for long stretches of time.

The appeal for me is that I can deliver a lot of water, in very little time.  Stick a hose into the core, run the rain barrel pump for 12 minutes, and I’ll have delivered the equivalent of more than an inch of rain to the bed.  I’ll need to do that, at most, once or twice a week if there’s no rainfall.  And I can do all that without putting any water on the surface of the bed.

There’s some up-front labor, though.  This requires digging a trench down the length of the raised bed, then filling that trench with straw (or similar).  Once that’s covered with dirt, the theory is that the straw “core” will act as a sponge, retaining water and slowly releasing it into the surrounding dirt.

With enough rainfall, a core bed waters itself.  Effectively, it’s a form of self-watering planter or sub-irrigated planter.  You store water in the base of your garden bed.  It keeps the soil moist, and you end up watering your plants from below, instead of above.  Rains periodically recharge the core.

That’s the theory, anyway.

I’m adding a little twist to this, in the form of some plumbing.  I want to be able to recharge that core from my rain barrels.  I have bits and pieces of PVC pipe sitting around doing nothing.  I’m going to se this up so that I can stick a hose right down into the core and rapidly refill it.

In my situation, this is more expensive than installing traditional piped-in irrigation.  That’s because I have to buy the straw for the core.  The recommended core size is foot wide and half a foot deep, running the length of the bed.  Doing the math, I need about 8 cubic feet of straw (or similar) for a small 4′ x 16′ raised bed.

In theory, that’s a couple of standard “two string” bales.  In practice, I only used one bale, despite doing my best to pack it down.   A standard two-string bale is (12″ x 16″ x 36″ = 4 cubic feet), but apparently expands by a factor of two no matter what you do to keep it packed.  A bale of straw costs $13 at my local garden center.  (That’s  enough to make no-dig gardening a value-destroying enterprise if you plant potatoes (Post #1703).  So each such bed requires about $13 worth of straw, per year, at the prices I’m paying.  That’s easily more expensive than a simple drip irrigation kit.

Obviously, YMMV, depending on the price of straw and the availability of substitutes for straw.

The upshot is that this isn’t for everybody, and it’s easy to see why this technique is not used commercially.  The MI Gardener uses it because he only visits his garden every couple of weeks.  I’m using it because I’m trying to use water barrels, at ground level, to water raised beds, at ground level.

The main upside for me is that it should greatly reduce watering chores.  I should have to water less, and when I do water, it should take almost no effort.


This is not rocket science.  Dig a trench, about a foot wide, and maybe 16″ deep.  Lay about 6″ of straw in the trench.  Mash it down, and cover it back up with dirt.  Maybe mound up the dirt a bit, as the straw will settle over time.  Then “charge” the core, that is, fill it with water for the first time.

I used one standard two-string hay bale for 16′ of trench.  (That’s a 12″ x 16″ x 36″ bale).  That provided the half-square-foot core cross section for the core (e.g., straw laid half a foot deep, in a foot-wide trench.) Those dimensions — foot wide, half a foot deep — seem to be the ones most widely recommended.  (The volume of the core of of this bed actually equals the volume of two bales, but apparently the straw must have expanded two-fold despite my best efforts to keep it compressed.)

There’s no barrier to using materials other than straw, as long as they’ll retain moisture well.  But straw seems to be what’s most commonly recommended, so that’s what I used.

I added one twist, which was to put a piece of PVC pipe into the core, so that I could water the core directly.  I considered by rejected the idea of running pipe in the core itself, to help distribute the water.

1:  The bed.  This is 4′ x 16′, with maybe a foot of dirt in the bed.  I put these together last year out of used campaign signs, bamboo poles, and metal fence posts (Post #G05, Wordless Workshop).

2:  The trench.  The trench should be at least a foot wide.  In this case, it goes all the way down to the red clay soil under the raised bed.  Only at the end, when I started to fill the core with water, did I realize that the trench bottom really should be as level as you can make it.


3:  Place the straw.  Aim for a half-square-foot cross section.  I ended up making the trench a bit wider than 12″ and the straw a bit less deep than 6″.  But it’s close.  Pack it down as much as possible — that’s why I’m standing on it.

4:  Re-cover the trench.  One thing I didn’t anticipate was that I didn’t have enough room for the dirt.  I ended up doing this in sections, burying the straw as I went. The white thing is a piece of PVC pipe that I had sitting around.  I set the end about halfway through the thickness of the straw layer, the piled dirt around to set up upright.

5:  Fill the core with water.  This is what the pipe is for, in this build.  Otherwise, just expose a piece of the core and run a hose to that.

This last step was essentially the proof-of-concept.  I put just under 50 gallons of water into the core, using water barrels and a pump.  At that point, water began to seep out the bottoms of the sides of the bed. An inch of rain on this 4′ x 16′ bed is about 40 gallons.  So I can, in fact, stick a hose into the dry core and quickly provide more the equivalent of an inch of water to that bed.

I will eventually replace that pipe with something less intrusive, and add a layer of mulch on top.  In the end, I probably don’t have quite enough dirt in this bed for this to work exactly right.  Even after mounding the dirt in the middle, that dirt layer is under the 10″ or so that seems to be recommended.  I may need to add some topsoil to get this to the point where I can grow plants uniformly across the surface.  But for now, I can just plant on the sides and not the middle.

A few questions and answers.

One of the things I dislike about gardening is that it’s hard to get advice that’s based on rigorous testing.  A lot of advice is something that somebody heard, or tried, and it seemed to work for them.  They pass that on.  With no actual controlled trial of it.  Last year, I found a lot of seemingly reputable gardeners, on the internet, who swore by advice that simply did not work, at all, for me.  (Post #G20, powdery mildew, search “seemingly reputable gardeners”).

In other words, much gardening advice is folklore.  It lacks a sound evidence basis.  It hasn’t been rigorously tested.  And, unfortunately, the only way to separate true from false folklore is to try it and see if it seems to work.

There’s a reason that it’s mostly folklore:  Almost all rigorous testing in agriculture is geared toward commercial farming.  That’s where the money is.  But if it’s strictly for the home gardener, chances are nobody will have done that expensive, large-scale controlled testing.  There’s just no money in it.

The bottom line is that I made this core bed because I saw a guy on YouTube who said it worked for him.  That’s the evidence basis for doing this.  Nowhere did I find (e.g.) a side-by-side comparison of a core bed and a regular mulched raised bed.  And so, I might just do that myself, and see if there is a noticeable difference in the need for watering.

I had a long list of questions I would have loved to have answered before I dug a trench down the middle of my raised bed.   Most of which I could not answer, but a few of which I could.  So let me document those.

Everything below here is theory, not practice.

Question 1:  How much water should the straw in this core be able to hold onto?

Answer:  About as much as I’d get from a third of an inch of rain.

Most kinds of straw will retain between two and two-and-a-half times their weight in water.  That’s based on this reference (via Google, page 188), or this one (page 19).  The core for this bed used one standard two-string bale of straw.  A two-string square bale weighs in at about 40 to 60 pounds.  This means that the core, for this 4′ x 16′ bed, should hold (50 lbs per bale x 1 bale x 2.1 water retention factor / 8.5 lbs per gallon) = 12.5 gallons, or about one-third of an inch of water, if spread over the 4′ x 16′ bed.

I was actually able to put almost 50 gallons into the dry core.  Presumably, most of the rest of that is in the spaces between the straw, and will run out into the surrounding soil quite rapidly.  Which, because this sits on almost-impermeable Virginia clay, means that it’s running out into the soil in the bed.  Which is just what I want.

Most vegetables that I grow seem to want an inch or two of water per week, in peak season.  That means I can probably get by with watering this bed once a week, even in hot weather with no rain.

Question 2:  Would lava rock work, instead of straw, for retaining water?

Answer:  Eh, maybe, based on the simple experiment reported here.

The idea here was, wouldn’t it be nice to use something permanent, such as  lava rock in sacks, instead of straw.  A secondary factor is that you never know what pesticide or herbicide residues are on the commercially-purchased straw.  Lava rock would in some sense be both cleaner and more permanent than straw.

It appears that lava rock (“scoria”) can retain water equal to perhaps one-fifth of its, per the experiment reported above.  That means that the 8-cubic-foot core of this bed would retain (8 cubic feet x 20% x 7.45 gallons per cubic foot =) 12 gallons of water, or just about the same amount as I estimate that the 50 pound straw core will retain.

On the other hand, lava rock is much more expensive.  Looking at the Home Depot website, 8 cubic feet of lava rock would cost about $80.  So, again, you aren’t going to see this used commercially.

On net, I’m not ready to do one with lava rock yet.  I’m not sure how well it would release water into the surrounding soil.  And, at this point, I think I’ve done the hard part.  The straw may rot, but the trench will still be there next year when I redo this.  I’m sticking with straw for now.

Question 3:  How about that polymer gel water-holding stuff?

Answer:  I think those “water storing crystals” are ruled out for bulk use because they change size too much as they gain and release water.  If I filled the core with that, the soil above it would be moving up and down as the gel got wetter or dryer.  I note, however, that you can now buy big pads filled with that, to use as a growing aid.  So, apparently, it is used, in bulk, as a way to store water in beds.

Question 4:  What about corrugated cardboard core?

Answer:  In terms of storing water, that would retain substantially more water than straw does.

People certainly use corrugated cardboard as mulch.  That said, I could not find information on how much water a piece of saturated corrugated cardboard would hold.

It was easy enough to measure, using a kitchen scale.  I took a piece of corrugated cardboard, weighed it, soaked it thoroughly, let it stop dripping, and weighed it two ways:  Once, held vertically (so all the interior channels drained), and once, horizontally (so they could retain some liquid water.  The cardboard held 1.2 times (vertical channels) to 1.5 times (horizontal channels) its weight in water.

A cubic foot of corrugated weighs about 40 pounds (reference). So an 8-cubic foot core, made of corrugated cardboard, stacked with channels vertical would retain (8 cubic feet x 40 lbs per cubic foot x 1.2 water-retention factor / 8.5 pounds per gallon =) 45 gallons.  If stacked flat, it would retain 56 gallons.

If it would absorb and release that water well, that’s massively better than a straw core.  I think that owes to the greater density.  Corrugated cardboard seems to be about three times as dense as straw.

But that’s a big if.  Plausibly, one of the advantages of straw is that it is so open, so that there is good transmission of water into and out of it.  And since my goal is to be able to recharge this core quickly, using a hose fed directly into the core, I think I’m not going to stick with straw.

Certainly, if I used cardboard, I’d have to run a pipe through the middle of the core to distribute the water.  So, plausibly, corrugated cardboard built around (say) a length of 4″ slotted PVC drainage pipe might work.  It also might work to tear up the cardboard and pack the pieces in randomly to provide space for water to permeate the core.

I can’t quite figure out whether this would have any environmental benefit or not.  On the one hand, cardboard is recyclable.  Putting out out to rot wastes the energy that went into the original manufacturing process.  On the other hand, hay bales have to be shipped, including that energy-intensive step of driving to the store and back to pick them up.  Either way, any benefit probably amounts to less than rounding error for me, given that I live in a home in the ‘burbs.

I have other questions, but I don’t see any way to get answers to them.  Mostly, I’d like to see:

  • A side-by-side comparison of the performance of cored and un-cored raised beds.
  • Results of some experiments with different core sizes to see what works best.
  • Some estimate of how common it is to buy straw bales with significant pesticide or herbicide contamination.


So far, so good.  This cost me $13 and an hour of labor.  It appears that the results are going to save me many hours of carrying watering cans around this summer.  Arguably, the continuous moisture provided by the straw core is better for the plants then periodic dry/wet cycles from watering with watering cans or hoses.

In theory, it should mostly refill itself from rain water. It more-or-less converts your raised bed into a self-watering planter.

It also has an appealing simplicity.  No plastic to degrade, no fixed irrigation points to work around, no pipes or hoses to trip over, no timers to break, and so on.  It’s just a trench full of organic matter, in the middle of your raised bed.  With a little more subtlety on the plumbing portion of this, no one would even guess that this is any different from a standard raised bed.

Assuming this all works out as planned, I’d do this again.