My DIY substrate circulation system designs

I've seen a lot of very interesting and insightful conversation based upon my 
ideas for a way to prevent anaerobic spots in the substrate and (by 
coincidence) also replenish the trace nutrients in without having to add more 
laterite to a tank.

I'm working based on the following information for my design...

1.  In order to prevent anaerobic decay in the substrate and H2S in the tank, 
you want a slow flow of aerated water to from the tank (or sump), through the 
substrate, and back into the tank (or sump).  

2.  I don't need to add extra heat to my tank because I run it at 84o F.  

After designing my setup, I realized that I could also use it to directly 
inject Duplaplant 24 drops into the substrate.  I understand that some of the 
Fe++ will escape into the water, but that's okay, the plants can use it there 
too.  The big factor is, my laterite stays loaded with more iron cations than 
it would if I just added drops to the water.  Since my design is open to the 
air at the top of the "fill" tube, I can add liquid and fully soluble solid 
fertilizers to the substrate directly instead of adding them to the water where 
algae have as much a shot at them as plants.

I recognize that my design has the potential for uneven flow.  I'm pondering a 
modification that might reduce this potential.  I also realize that my design 
gets the Fe++ out of the light quickly so it has fewer oxidation factors 
working against it.

I guess I should describe the design for those of you that haven't seen it.

3-D BUAG follows:
     /                                                 /
   /                                                 /
 /                                                 /

| = 1/2" PVC Fill tube (clear or otherwise)
/ = 1/2" PVC "manifolds" for distributing the water to the runners
= = 1/2" PVC runners, perforated with 2 rows of small holes along the
    underside.  The holes may need to increase in size as distance from
    the fill tube increases.

An anti-siphon hole in the return line from the sump sprays water into the fill 
tube.  The fill tube extends above the water level and is open at the top so 
any excess water spills over into the tank.  The head pressure of the water 
column above the water level in the tank forces water out through the holes in 
the runners.  This is a RUGF-type system, it just has a very slow flow.  My 
design doesn't start with any media of any type inside any part of this system. 
 I'm very interested in the ideas that are surfacing, though.

My altered design that may help with the uneven flow problem is as follows: 

     /                                                 @
    /===============================================/ /
                                                   / /
  /===============================================/ /
 /                                                 /

@ = a minijet pump to circulate water inside the system.  

I'm not sure that this will actually solve the uneven flow problem, or just 
move the location of the problem inside the system.  I'm still pondering it.  
If the substrate flow pump breaks, I have no way to fix it without digging up 
my system.

Note:  I haven't implemented this system yet.  I've mentioned my ideas and 
people have asked about them, so I've shared them.  I'm very gratified that so 
many people find them worthy of mention.

I do plan to implement a version of this system in the near future.  I have one 
box of Duplarite G that I will probably mix with boiled vermiculite and sand to 
make up the lower substrate.  The upper substrate will be a fine or medium 

This is only my design for a substrate flow system.  My other designs have been 
in use for years in the case of the "smart siphon" or work very well in theory 
in the case of the "automatic vacuum".  In fact, they are only slight 
modifications of one another.  The "smart siphon" is the only one of these that 
doesn't require at least a partial teardown to implement, and I haven't torn my 
tank down yet.  I am very interested in all opinions on the substrate flow 
design, even negative ones, but I currently believe it will work nicely.

Please pardon me for wasting so much bandwidth, but I thought I should explain 
my design for those who weren't familiar with it.

David W. Webb
Enterprise Computing Provisioning
Texas Instruments Inc.
(214) 575-3443 (voice)		MSGID:		DAWB
(214) 575-4853 (fax)			Internet:	dwebb at ti_com
(214) 581-2380 (pager)		Text Pager:	pgr at msg_ti.com Subj:PAGE:David