A Guide to Back-Flood Swales
This article talks about some of the design issues you’ll face when constructing a back-flooding swale, the signature of Mr Geoff “Reconstructive Earth Surgeon” Lawton.
It’s a great idea and provides a few additional beneficial functions to a standard valley dam, namely increasing the catchment by whatever length the contour trench wraps around the landscape, as well as utilising any dam overflow quite effectively by spreading it around the landscape and infiltrating it into the soil reserves.
However, water’s erosive potential must be respected and hopefully, as well as making it easier and less daunting for people implementing Earthworks for the first time, my aim in writing this article is to help them avoid some potentially embarrassing, destructive and very expensive mistakes.
Here’s my attempt to perpetuate the Permaculture water infiltration nomenclature confusion. To those from the US who read this site, I do realise that Bill gave us all a bum steer and the true definition of a swale is a sloping drain. Here in Australia we usually call a spade a spade (or a drain a drain), but in this case, seeing as though no one here calls a drain a swale, therefore we are able to call what we call a swale a swale and get away with it. Clear as mud? Hope not, cause if so your back-flooding swale might be a US swale drain not a non-draining Aussie swale! PS in the retaliation of US/Aussie definition wars, if you build a wall of earth in a valley to hold back water, that is by definition a dam, not a pond. Cheers :-)
1. Respecting Traditional Dam Design
Living on the driest continent on Earth, Australians soon become pretty good at capturing and storing rain water. As a result you could safely say that there is a small earth dam on almost every rural property around the country.
Water can be a very destructive force and to avoid the costly loss of a dam wall (including the often irreplaceable earth construction material) some tried and tested conventions have evolved (See Design and construction of small earth dams by K Nelson for an excellent resource).
One of these conventions is the freeboard of a dam (Figure 1.1), which is the height difference between the (settled) top of a dam wall and the level of the spillway (which sets the height of the water in a full dam).
Freeboard: the height difference between the spillway
and the top of the dam wall.
In conjunction with a spillway that has the capacity to handle the flood flows received from the associated catchment during a 1 in 100 year rainfall event (which could come at any time of course), the freeboard prevents water from flowing over the dam wall and eroding it away. For small earth dams, this is generally 1m, which allows 0.5m of flow over the spillway and 0.5m for wave action (can be less of course for a small catchment as long as the spillway is wide enough). The appropriate sizing of spillways and freeboard can be obtained from the local catchment management authority or from K Nelson’s book I mentioned above.
When an existing valley dam is retrofitted with an adjoining swale, or a new dam-swale combo is constructed, the recommended dam freeboard must be maintained to protect your dam wall (a large cash investment). Hence, the existing freeboard sets the height of the water in the back-flooding swale (Figure 1.2), which in turn is set by the height of the chosen spillway.
2. Swale/Dam Connection Options
There are a number of options available to us when designing a swale that’s connected to a dam. Here are some diagrams of a few possibilities, with an exaggerated swale water depth used to illustrate the effect of each option.
Option 1. The Open Ended Back-flooding Swale
This is the option that is generally referred to when a back-flooding swale is described:
Figure 2.1a: As shown above, the end of the swale that’s attached to the dam
is open, and since as Geoff points out “Water can’t stand on its head”,
any decent runoff that enters the swale spills out into the dam.
Figure 2.1b: Once the level of the dam reaches the base of the swale,
the dam and swale rise as one body of water, overflowing from the spillway.
Figure 2.1c: Because the swale mound is uncompacted and is designed
to infiltrate, the water in the trench soaks in, and so does the water in the
top section of the dam.
Therefore, if the recommended freeboard (ie 1m) is respected and the chosen depth of water in the swale is 40cm, following infiltration into the swale, the resting height of the dam will be 1.4m below the dam wall. For every 2500 m2 of surface area, this 40cm of reduced dam storage represents either 1 mega litre less storage capacity in the dam, or 1 mega litre of increased infiltration into soil stores, depending on the land manager’s water use preferences and priorities. (ie less water available for stock troughs and drip irrigation to young trees for which deeper water storage won’t be as helpful)
The other downside of this design option is that until the dam fills, very little water is infiltrated into the swale mound, so groundwater to the trees below is effectively drained rather than being hydrated.
Option 2. The Dividing Lip Back-flooding Swale
This second option consists of a lip of undisturbed earth between the swale and the dam, which is set at about 100mm below the height of the spillway.
Figure 2.2a. When runoff enters the swale, the water remains in the trench
until it reaches the height of the lip. At that point, the water begins
to spill over and fill the dam
Figure 2.2b. Once the dam is full, the dam and swale will rise as one body for the
remaining 100mm until the spillway height is reached.
Figure 2.2c. Following a large rain event, due to the lip of undisturbed
soil the dam will hold its level, whilst the swale will infiltrate
This option solves the potential problems of the reduced dam storage and less frequent swale infiltration in option 1, however the downside is that the dam catchment is only increased when the swale is already full.
Option 3. The Pipe Connected Back-flooding Swale
This option includes a culvert pipe that rests on the floor of the swale, connecting the dam and swale. In my opinion it offers greater control to the land manager over how the water behaves. (A couple more pipe options will be discussed later on). Here are the basics:
Figure 2.3a. If filling the dam is the priority, the pipe is kept open and
the system behaves like option 1.
Figure 2.3b. Alternatively, if hydrating trees below the swale is the priority,
the pipe is blocked and the system behaves like option 2.
Figure 2.3c. When the dam reaches the level of the base of the swale,
they rise as one body of water and flow over the spillway.
Figure 2.3d. Various simple ways of blocking the pipe allow the
full depth of water to be maintained in the dam.
Figure 2.3e. If the dam is full and the swale empty, as in 2.3d, the pipe
can be opened to allow the dam to fill up the swale if required.
3. Design Issues
This article is more about the finer details of a back-flooding swale rather going into all of the design issues related to the siting, sizing, and spacing of multiple swales. (See Rainwater Harvesting Volume 2 by Brad Lancaster for more detail on that topic.)
However, here are a few points to consider:
- The placement of a back-flooding swale will often be set by the position of an existing dam.
- If the swale and dam are constructed at the same time, placement will usually be decided by:
– the most efficient and economic placement of the dam in relation to the shape of the valley in which it is placed, or
– the placement of the dam in a slightly less efficient position to take advantage of the increased catchment provided by a special source of water that the swale can pick up, such as another valley further round the landscape or a road culvert.
- The size and spacing of a swale is a balance between a number of factors, including climate, rainfall intensity, soil type, catchment vegetation cover, and the associated tree and farming practices. Some of these are discussed in more detail in Brad’s book that I mentioned above. However, the balance of two calculations are integral and can be helpful to mention here:
– Determining the volume of water that will flow into your dam & swale (it’s worth considering their catchments separately and as a whole) during a large rainfall event:
Run-off area (m2) x large rainfall event (mm) x estimated runoff (%)/100 = runoff volume in litres
- Determining the volume that your dam and swale will hold (once again, it’s worth considering them separately and as a whole).
Water cross section (see 3.1 below) of full swale (m2) x length of swale (m) x 1000 (litres) = swale volume in litres
Figure 3.1 Water cross section of swale
Determine soil character
- Always test whether soils are dispersive, or else tunnel erosion is quite possible
- Swales are not suitable in slip country
- Heavy soils may require ripping or gypsum to encourage infiltration
- Salinity: Although there are examples of the water infiltrated by swales creating a freshwater lens that perches above the salty groundwater (such as in the flat Jordan landscape in Geoff’s Greening the desert project), care must be taken in sloping country where salinity is an issue:
– The increased infiltration caused by a swale could contribute to salt outbreaks further down the slope, particularly while tree systems are small and not utilising the soil moisture.
– Smaller banks more often in the landscape will provide less downward hydraulic pressure on the saline groundwater than a couple of large trenches.
– Avoid cutting deep into the subsoil in saline areas. Aim instead for moisture to soak predominantly into the A horizon, which may have more likelihood of creating a freshwater lens on sloping country.
Full Water Depth in Swale
- The depth of water in your swale when it’s full will be decided by the height of the spillway in relation to the base of the trench See Figure 3.2.
- When you are dealing with potentially large flows from the valley that enters the connected dam, the mound needs to have a sufficient freeboard above the spillway height to handle the increased water depth (although not as high as a dam because wave action won’t be as big a deal)
- Another factor which you have control over is the ratio of water that will rest below or above the mound cut (Figure 3.2), the latter soaking into the mound, the former into the original soil.
- I personally like to go for a mix of about 50:50 below:above to make sure my trees on the mound get enough moisture while young.
4. Constructing a Back-Flooding Swale
To have enough capacity to perform its functions, a back-flooding swale needs to be built with a decent sized machine such as a bulldozer or an excavator.
4.1 – Construction with a side casting bulldozer or 6 wheel grader
When using a bulldozer or grader to construct a swale, survey pegs are placed on the upper edge of the trench as a guide. It would be nice to be able to give a blanket statement for how deep the dozer should go in relation to these pegs, but unfortunately this isn’t possible. For example, in steeper country, the dozer can cut down to a blade’s depth below the survey peg and leave a suitable ‘mound cut’ height at the front (Figure 4a). However, as the land gets less steep in figures 4b and 4c, you can see that the blade begins to cut to ridiculous depths.
Figure 4a Steep country
Figure 4b Medium country
Figure 4c Gentle country
As a result, I came up with the instructions below, which are the closest I can come to communicating how to get a good finished result with a dozer in any shaped land. (Any feedback is more than welcome to help improve this info.)
4.1.1 Pre-Construction Survey (side-casting bulldozer)
1. Determine the spillway height for the system, based on maintaining an adequate freeboard in the connected dam (See section 1) and place a temporary height peg as a starting point (green).
2. Determine the desired water depth at high tide. Eg. 0.4m
3. Determine water depth below the mound cut (ie 0.2m) and above the mound cut (ie 0.2m). (See below and the explanation at Figure 3.2)
4. Measure an approximate average gradient for the selected site (the gradient will change as you move in and out of valleys and ridges, but the aim here is to measure what you think is an average slope)
5. Measure the dimensions of the bulldozer blade (or grader blade if it is flatter country).
6. Set the bulldozer guide pegs:
- Start at spillway height (green stake)
- Add average gradient x Blade width (e.g. 10% x 3m = 0.3m)
- Subtract Water depth above mound cut (See Figure 3.2), e.g. 0.2m
- Place first bulldozer guide peg (red stake), e.g. Green stake + 0.3m – 0.2m = Guide peg @ 0.1m higher
7. Using a laser or water level, work from the original bulldozer guide peg, banging (sacrificial) wooden stakes @ 10-20m intervals on contour for the desired swale distance. (Note, it helps to paint the pegs with a bright colour. Do this when the pegs are lying down before you put them out, it’s much quicker and uses less paint)
8. Determine width and position of spillway(s) (see section 5). For extreme spillway accuracy, put stakes at 1m intervals at the desired high water mark.
4.1.2 Construction (side-casting bulldozer)
- Get hold of the best operator in the region. Ask around.
- Spray paint the cut depth mark on the rear of the blade (clean it off for him later with a wire brush).
I. Start at base of blade
II. Add Gradient x Blade width eg + 0.3m
III. Add Water depth below mound cut eg + 0.2m
IV. Mark rear of blade @ appropriate height eg. = 0.5m
- Depending on soil conditions, the bulldozer rips and side casts until the depth mark on the blade intersects the soil directly beside the guide pegs. Important: avoid cutting the spillways until last. Leave these sections intact.
- Subsequent passes batter the rear wall (the driver will know the appropriate angle for the local soils). On steeper sections of the landscape, you will need to take off a bit more from the rear batter to get enough material for the mound
- Remaining passes are to trim spoil from rear batter and take off any high points when checked by laser level. Also check the mound for any low spots.
- Check spillway height carefully before proceeding. Spillways can be cut roughly by the bulldozer to within 200mm of the upper edge, but for better accuracy should ideally be finished with a post hole shovel
4.2 – Construction with a tilt bucket excavator
A tilt bucket excavator can be a useful tool when building a swale. It can travel backwards along the contour cutting the base of the swale and then battering the back edge while swivelling sideways to build the mound from the spoil.
For a tilt bucket excavator the survey pegs are placed at the lower edge of the trench. A standard excavator traverses the hill below where the mound will go and pulls the trench material downwards to build the mound. In that instance the survey pegs are placed at the top edge of where the batter will finish.
4.2.1 Surveying (tilt bucket excavator)
- Follow the same first three steps for the bulldozer survey above, deciding on:
– the spillway height
– the depth of water when the swale is full, and
– the depth of water desired both above and below the mound cut at the front edge of the swale.
- To place the first excavator guide peg:
– Start at the spillway height (green stake)
– Measure vertically down the depth of water you want the swale to hold above the mound cut (see Figure 3.2)
– Place your first excavator guide peg (red stake)
- Using a laser or water level, work from the original excavator guide peg, banging stakes @ 10-20m intervals on contour for the desired swale distance.
- Determine the width and position of spillway(s) (see section 5). For extreme spillway accuracy, put stakes at 1m intervals at the desired high water mark.
4.2.2 – Construction (tilt bucket excavator)
- Dig on the uphill side of the guide peg down to the chosen depth and place the material on the lower side. To keep the depth accurate, it’s good to have someone on the laser level working with the excavator operator.
- Tilt the bucket to batter the back bank
- Batter the front cut and mound to the desired shape
5. Spillway Siting and Design
5.1 An argument for retaining the traditional dam spillway
When valley dams are retrofitted with a back-flooding swale, often the original spillway is forgotten and the swale constructed over the top. The reason for this is that it’s presumed that a level sill spillway of the same width placed further around on a ridge will perform the same function. Where a large catchment is involved, such as 50-100Ha, here’s a couple of diagrams that suggest why this won’t work in a big event.
Figure 5.1.1 – A valley dam spillway during a 1 in 100 year event,
in this case with 0.5m of water flowing over a 10m wide spillway.
The spillway width is determined using calculations related to the
catchment area and the expected size of a 1 in 100 year event
(information available from the Catchment Management Authority).
Figure 5.1.2 – Does this stack up? Here is the same amount of water
that passed over the spillway above in relation to quite a large sized swale.
If the system’s only spillway was situated further around on a ridge, the swale would have the job of transporting this water around to the spillway. As you can see, in this case the swale will struggle to transport a flow towards the spillway of even half the expected major event.
What happens when you pour petrol too fast into a funnel? It fills up and spills down your leg. In this case, the place that water will spill if it can’t get through the swale is potentially over the dam wall. That is a disaster.
By incorporating the traditional spillway, this problem can be avoided
5.2 Designing to incorporate the traditional spillway
A trench cut in behind the traditional spillway allows the dam and swale to be linked. The trench could take the form of any of the three options mentioned in section 2 depending on your management objectives. Figure 5.2 explains the rest.
5.3 General ‘level sill’ design considerations
When designing a level sill spillway, always follow the same rules as for a dam spillway, that is:
- The height of the spillway determines how high the water will fill in the dam and swale.
- Assuming the traditional spillway is sized appropriately for the valley catchment above, the same needs to be done for the catchment area directly entering the swale (same as for local dam planning). If multiple valleys enter, build a spillway for each on the adjacent ridgeline.
- Make it dead level
- Always construct it on original ground
- Keep the spillway grassed and clear of debris
- Where possible, it’s best to spill water on a ridge, that way it fans out with even less chance of causing erosion (see figure 6.4)
- As seen in Figure 3.2, the spillway height is usually higher than the mound cut at the front edge of the swale trench. Seeing as though we want to spill the water over original ground, this means we have to take our pegs uphill a bit when constructing the level sill (Figure 5.3.1)
Figure 5.3.1 – Marking out a level sill spillway
- As seen in figure 5.3.2, it helps to put a little extra ‘uphill kink’ either side of the spillway, which being undisturbed soil will help to protect the mound.
- The mound needs to wrap in a little on either side of the spillway or else the water will concentrate and flow out the side of the spillway at the height of the mound cut. It’s also important to wrap the mound around at the end of a swale if there isn’t a spillway placed in that position.
Figure 5.3.2 – Plan view of a level sill spillway
6. Dissipating Large Flows
Figure 6.1 – Within the bounds of the property, neither valley lends itself
to an efficient place for a dam (generally the minimum backup worth
constructing is equal to the width of the dam wall). Hence a swale has been
placed to make use of the flow entering from above.
Figure 6.2 – During a large rain event, there is potential for the flow
to break through the swale
Figure 6.3 – A small dam/wetland pushed up acts to dissipate the water’s
energy before bleeding it sideways into the swale. Wetlands also acts as a
sediment trap, as well as distributing dissolved nutrient along the swale.
It could also provide gravity fed water to stock if large enough and will
of course provide habitat.
Figure 6.4 – Spillways placed on ridgelines encourage the water to fan out,
giving it more time to hydrate the ridges below (occasional keyline plowing
will help further). The red arrows indicate emergency spillways in times of
very large flows (placed slightly higher than ridge spillways)
7. Silt Traps
Occasional deeper sections within the swale can act as a silt trap. This is where Peter Andrews might dump a dead horse to fertigate the slope below during the next few rain events.
8. Crossing Design
A big, long swale can create quite a barrier and restrict access to sections of the property. Therefore crossings may be necessary for vehicles, stock or even just footpaths on a small scale; the principles are the same.
Figure 8.1 – Plan view of a standard swale
Figure 8.2 – Culvert pipes laid in base of trench
Figure 8.3 – Mound wrapped in at each end of pipes, level from
the top of the swale mound until it reaches the slope behind.
Alternatively a rock wall can be built either side.
Figure 8.4 – Road or path filled in, graded and top-dressed
Note: Make sure you have a spillway either side of any culvert pipe, just in case there is ever a blockage
9. Pipes for Control
The versatility and control that can be gained by incorporating culvert pipes in strategic positions was illustrated in Section 2, Option 3 – The pipe connected backflooding swale.
If water-logging of soils is a possibility (such as during winter in a temperate environment with heavy soils and dormant deciduous trees coinciding with a winter dominant rainfall pattern) or excessive leaching of nutrient due to very high rainfall in sandy soils, another useful pipe addition is a culvert placed under the swale mound (Figure 9.1).
Figure 9.1 – Culvert pipe under swale mound
Figure 9.2 – Pipe open: helpful during wetter periods where excess water
is a potential liability due to waterlogging or excessive leaching
of nutrients in the dormant season
Figure 9.3 – Pipe blocked: Helpful during dry periods to catch
and infiltrate any available runoff.
A few options for blocking a culvert include:
- a piece of marine ply with a rubber backing, held in place by two star pickets
- a sheet of heavy tarp, a la the flag used in the Keyline flood irrigation system
- a slightly deflated basketball wedged in
- an appropriately sized plastic plant pot with sloping sides, filled with concrete, with an inner tube wired around the outside to plug the hole
On a smaller scale you can do similar things by linking garden ponds with hidden soakages under pathways. The wide range of PVC pipe fittings available allow you to do so with even more options, more water control and can require less management than the larger culvert pipes in the article. (See www.forestedgepermaculture.blogspot.com for a couple of articles I wrote a while ago.)
Cam Wilson runs Earth Integral, offering design, implementation and education in landscape rehydration and rehabilitation in the Southern Tablelands, NSW, Australia