I have seen at various time the flow velocities of 1 to 2 ft/second as the minimum
value for flushing driplines. I do not have any reason to dispute this value, but I would
like to find the original source of this information or alternatively maybe another body
of literature on which it based (Ie. wastewater systems). Can anyone help me out.
I have already looked at
which gives a good discussion but no indication of original source.
The following discussion may help.
TURBULENT FLOW AND LINE FLUSHING 12/20/99
by Dr. Alvaro Sanjines.
The purpose of this discussion is to determine when is there enough velocity in the
pipe to carry over and properly "flush" driplines.
Flow regimes are characterized by the Reynolds number, a non-dimensional factor that
was empirically found correlates with the flow pattern.
During laminar flow, at lower speeds, flow streams are parallel to each other and
gravity will settle particles at the bottom of the pipe. As speed increases and hence
Reynolds number, eddies and turbulence appear that keep particles suspended and in
motion, helping flush the line properly.
The Turbulent flow regime is defined by the Reynolds number
Nr = Velocity (ft/sec) * Diameter (ft) / kinematic viscosity (ft2/sec)
The kinematic viscosity varies with temperature
at 60 F is 0.0000121 ft2/sec
at 80 F is 0.0000093 ft2/sec
at 100 F is 0.0000074 ft2/sec
To express this equation in an easier form:
Nr = 3277 * Flow in GPM / Pipe Diameter (inches)
This value is for ambient temperature of 70 F.
Now, the laminar region is for Nr less or equal to 2,000 The
turbulent flow region is for Nr larger than 2,500. The question then is how much larger
than 2,500 do you want the Reynolds number to be. Higher pressures and larger pumps are
required in order to flush the lines if the standard is set at higher Reynolds numbers. A
Reynolds number of 4,000 seems like a reasonable choice to assure turbulence in the lines.
For various Reynolds numbers, a Table was calculated, giving the flow that has to exist
at the end of the pipe to assure that turbulent flow exists all along the pipe. For
standard WASTEFLOW?which has an internal diameter of 0.52" this requires 0.63 gpm
or 0.98 ft/sec per lateral.
The Reynolds number will be higher at the entrance and middle sections of the pipe
where the flow is always larger as there is leakage through the emitters along the pipe,
when flushing. The values for required flow at the end of the line to attain Turbulent
flow are calculated.
A model can be set up to calculate what the pressure should be at the beginning of the
line to assure turbulent flow at the end of the line, taking into account the leakage at
each emitter along the dripline. This is also dependent on the length and diameter of
DISTAL FLOW IN GPM FOR VARIOUS REYNOLDS NUMBERS
70 Degrees F Internal Diameter (Inches)
Reynolds 0.5 0.52 0.54 0.56 0.6 0.7 0.83
2000 0.31 0.32 0.33 0.34 0.37 0.43 0.51
2500 0.38 0.40 0.41 0.43 0.46 0.53 0.63
3000 0.46 0.48 0.49 0.51 0.55 0.64 0.76
4000 0.61 0.63 0.66 0.68 0.73 0.85 1.01
7000 1.07 1.11 1.15 1.20 1.28 1.50 1.77
8000 1.22 1.27 1.32 1.37 1.46 1.71 2.03
9000 1.37 1.43 1.48 1.54 1.65 1.92 2.28
10000 1.53 1.59 1.65 1.71 1.83 2.14 2.53
DISTAL END VELOCITY IN FT /SEC
70 Degrees F Internal Diameter (Inches)
Reynolds 0.5 0.52 0.54 0.56 0.6 0.7 0.83
2000 0.51 0.49 0.47 0.45 0.42 0.36 0.31
2500 0.64 0.61 0.59 0.57 0.53 0.45 0.38
3000 0.76 0.73 0.71 0.68 0.64 0.55 0.46
4000 1.02 0.98 0.94 0.91 0.85 0.73 0.61
5000 1.27 1.22 1.18 1.14 1.06 0.91 0.77
6000 1.53 1.47 1.41 1.36 1.27 1.09 0.92
7000 1.78 1.71 1.65 1.59 1.48 1.27 1.07
8000 2.04 1.96 1.88 1.82 1.70 1.45 1.23
9000 2.29 2.20 2.12 2.04 1.91 1.64 1.38
10000 2.54 2.45 2.36 2.27 2.12 1.82 1.53
This information was contributed by Rodney Ruskin.
I ran into the same problem some years ago. Most references at that time indicated
flushing velocities of 1 fps. We know that 1 fps is also often used as the maximum
velocity for settling basins, so my thinking was that it was insufficient. So our
recommendation was to double it - 2 fps. But I never found a good reference based on
research to support that rate.
The problem with 2 fps is that with some standard designs and pumping capacities
(pressures)it may not be achievable.I would like to hear a good answer to your question.
Has anyone written a metric version of this? It's double dutch to us Brits these days!
In my experience, this approach (using a Reynold's number) isn't
correct. If you notice the values in the tables, as the pipe diameter increases,
the critical velocity decreases (such is the nature of Re number). If you just think about
it, in canals we have rules of thumb such as "maintain the velocity above 2 ft/sec
(.6 m/sec)to avoid sedimentation", and the scouring velocities are higher. Also, if
you extrapolate the tables, you see extremely low velocities for larger pipes - something
we just know isn't correct with the Re number approach.
There has been a fair amount of work on scouring and sedimentation, but unfortunately
the results are inconsistent. Therefore, it's difficult to give precise velocity
I appreciate all the leads about possible origins of the 1-2
ft/second dripline flushing velocity. Additionally some have sent private messages, which
I hope to summarize at some point in time for the group.
Here is a little further clarification to my question that would help me track the
Has anyone specifically seen in microirrigation publications that the origin of the 1 -
2 ft/second is credited back to some specific source or body of scientific
literature??????? IF SO, could you point me to the microirrigation publication AND the
Thanks again. I'll try to summarize what I find out. Once again, I am not disputing the
values at this point, I just want the origin and rationale for reference.
A little work with the math will quickly show the very significant
advantage of using pressure compensating drippers when a defined flush velocity is
When flushing, relatively little flow will be going throught the emitters relative to
the flushing flow, so I don't see the advantage of pressure compensating emitters in this
instance. Maybe I'm missing something??
Fred H. Harned
I have a little trouble with the last two statements from Fred. In
North Carolina to the best of my information the facts are as follows:
The body involved in this matter is the NC Department of Environmental Health and
Natural Resources. The application involved is the disposal of treated human effluent
through buried drip systems.
There are two main suppliers of drip tube for this application in NC. Company A has a
major client who in turn has a patent which includes the words "forceful turbulent
flow" to scour the inside walls of the tube. In general these clients of Company B
appear to teach a minimum velocity of 2 feet per second. Company B has a patent on an
interior lining of the tube which incorporates a bactericide and Company B insists that
the dripline only needs to be flushed to remove accumulated fine particles and that
scouring of the tube is unnecessary - i.e. 1 foot per second is more than enough.
Clearly Company A's method requires larger pumps than does Company B. Therefore, in my
personal opinion, the argument in NC has a lot more to do with the commercial interests of
Companies A and B than the engineering information which is the center of this discussion.
On the second point:
Company A offers only pressure compensating drippers into the sewage market. Company B
offers both turbulent flow and pressure compensating drippers into the sewage market.
Following Company A's concept:
When flushing at 2 ft. / second the flow from the pressure compensating drippers will
remain uniform and all the extra energy and flow will pass through the tube. Good for
scouring the tube but no benefit on cleaning the drippers. This is as claimed by Fred.
When flushing at 2 ft. / second the pressure and hence the flow from the turbulent flow
drippers will usually increase and some of the extra energy and flow will pass through the
drippers. Good for cleaning the drippers but to maintain 2 feet per second to scour the
tube even larger pumps are required. But, as pointed out above, Company A does not offer
this alternative and Company B does not teach the need for the higher velocity flush.
Following Company B's concept which just involves opening the end of the tube to flush:
Under virtually all practical circumstances the flow from the pressure compensating
drippers will remain uniform and all the extra flow will pass through the tube. Good for
scouring the tube but no benefit on cleaning the drippers.
On the other hand the flow from the turbulent flow drippers will drop and all this
reduction of flow will pass through the tube. Good for scouring the tube but no benefit on
cleaning the drippers.
Some civil engineers who are very risk averse will design systems combining the most
conservative parts of the methods taught by both Companies A and B. From my experience the
human effluent slime problem can more difficult to manage than the usual agricultural
bacterial slimes and hence requires stronger action - i.e. either high flush velocities as
taught by Company A or bactericide lined tube as taught by Company B.
Nothing is ever simple.
Dale Bucks and F.S. Nakayama at the 3rd. Int. Micro Irrigation
Congress, Fresno, CA Nov. 18 -21 1985 page 125: "A minimum flow velocity of 0.3
m/sec.(1ft./sec.) is needed for flushing lateral lines." Scott Tollefson at the 3rd.
Int. Micro Irrigation Congress, Fresno, CA Nov. 18 -21 1985 page 404: " The flush
manifolds are designed with a 3.5 kPa (0.5 psi) pressure loss .... "
Plugging of microirrigation systems is a major problem, and it may
occur from single or a combination of multiple factors. Physical factors, such as
suspended materials passing through filters or broken pipes, root intrusion, and
aspiration of soil particles into the emitter orifices are common causes of plugging.
Flushing velocities must be high enough (at least 2 feet per second) to transport and
discharge heavy particulate matter from the pipelines. Lateral lines should never be run
The real advantage occurs in the low pressure systems (e.g., tape)in
which you need to raise the pressure at the head end considerably to pass a flushing flow
rate through the end. The percentage hose inlet flow change (for flushing) is increased if
the initial emitter flows are low. But anyway, the high flushing pressure is a large
percentage increase, which results in a high percentage increase in flow out the emitters
along much of the hose. Exactly what happens depends upon the emitter exponent and the
pressure at the downstream end. In our book Drip/Micro Irrigation, we have a whole chapter
dealing with this. Our software also predicts the impact of whatever variables you want to
use during flushing.
Rodney, The engineering community that designs wastewater systems
should look at short and long term solutions. We weigh the risks and costs and make
decisions. One of the decision factors is how well can the system be maintained over its
expected life cycle. My understanding is that the company "B" approach is to
rely on a chemical barrier to protect it against potential plugging from fouling and
design it with a minimal capacity for flushing.
Company "A" as you have stated, relies only on scouring
for long term maintenance. Company "B"s' solution would appear to be a short
term solution since most chemical additives to plastics have a finite life, whether as a
liner, impregnated into the resin or applied as a layer. Leaching and oxidation eventually
use up or lower the potency of the chemical to where it no longer provides sufficient
protection. In the system that company "B" is proposing, the point at which the
chemical barrier is used up would appear to be the end of the system life cycle.
As engineers and designers, we need to know when that point in time
will occur, 10, 20 or 30 years? Above ground and in-tank components can be changed or
replaced as needed. Underground tubing has to do better than that. Chemical root barriers
used today have a finite life that I feel is too short as compared to the expected life of
a wastewater system, what about chemical barriers for bio-fouling? I'd like to see 25 to
30 years as a minimum. Energy savings and lower pump costs may not justify company
"B"s' solution in the long run.
My thought is that if using the Company B approach the flushing and
scouring would be designed into that system also. If not, why not?
It is designed in if the designer is competent. The question is what
velocity is needed to flush the hose, whether it is that produced by "Company A"
or "Company B". I believe, based upon my experience, that what is needed to
flush depends to some extent on what is put into the hose.
There is an approach to "waste" water management that
proposes to accomplish "disposal" (should we even be talking disposal instead of
reuse when drip irrigation is used?) by forcing septic tank effluent through physical
filters, then routing it into drip hose. That is the patented process using "Company
A" hose that Rodney referred to in his original discussion. This pretreatment process
does nothing to deal with bacteria and nutrients not bound up with solids above the size
removed by the physical filters, so there is a LOT of the slime producing agents routed
through the drip hose. Indeed, part of the patented process is a controller that runs a
flush cycle something like every 5 dosing cycles, because it is to be expected that stuff
will build up on the pipe walls (and be deposited on the pipe "floor" too) very
Another approach is to provide high quality pretreatment that
removes the vast majority of bacteria and much of the nutrient content from the
"waste" water before routing it to the drip irrigation field. (My designs use a
high performance biofiltration system for this treatment.) In this case, there would be a
much lower likelihood of significant slimes forming on hose walls, regardless of the
chemical additives that are or are not in the hose materials, and flushing is suggested as
preventative maintenance at much longer intervals. Recommendations run from 6 months to
2-year intervals. So, regardless of whether you use "Company A" or "Company
B", the maintenance liabilities are lower with the high quality pretreatment
approach, so maintaining a certain scouring velocity is much less critical. If the stuff
is not sufficiently thick that it won't shear off with a 1 fps velocity, it is also
unlikely that it will shear off during normal operation of the hose system. David
Venhuizen Just for the edification of the list, "Company A" is Netafim, and Mr.
Prochaska sells their tubing. "Company B" is Geoflow, and Mr. Rushkin is the
head of that company. This should give list members not already privy to this information
some perspective on their respective views. BTW, I did not read Mr. Rushkin's comments to
suggest that provisions should not be made for flushing. He was merely addressing the
flushing velocity required.
As I wrote in my other post today, Mr. Prochaska's perspective on
the need for flushing is undoubtably "colored" by what he considers to be the
"normal" type of system that uses his Netafim tubing. Thus he is rightly
concerned with assuring that the tube would be effectively scoured during flushing.
Just to add a further question to the issue. It would appear that
there is a significant difference between "flushing" & "scouring"
in drip line. As flushing involves the removal of sediments that have settled within the
drip line and scouring involves the removal of deposits that have attached to the wall of
the drip line, is the required velocity the same?
I would think that the velocity required to actually
"scour" the slime and attached deposits would need to be far greater than to
just "flush" sediment. I understood that the sales pitch of the manufacturers
with regard to chemical barriers added during the manufacturing process applied to root
intrusion barrier chemicals only. If any manufacturer is including chemical barriers to
slime or deposit build up I have many client that would be very interested.
My apologies to those of you that may be bored by this discussion.
However, by the numerous replies both publicly on Trickle-l and privately to me, it is
apparent that there are several lines of reasoning out there and they may not all be using
the same scientific basis. My original question was about the original source of the 0.3
m/s or 1 ft/sec flushing velocity.
I have confirmed through the help of others that the following paper
by Marvin Shearer indicated this value as early as 1977. He supposedly retired in the
1980s. Shearer, N.M. 1977. Minimum screening and automatic flushing. In proceedings of the
4th Annual Int. Drip Irrigation Association mtg, Fresno, CA. pp 32-36. Can anyone provide
me a copy or the name of someone who might???? Please email me directly to avoid