EPA
Water Sense
Preliminary Comments on the WaterSense®
Draft Specification for Soil Moisture-Based
Irrigation Control Technologies
January 23, 2020

-------
WaterSense
Preliminary Comments on the WaterSense
Draft Specification for Soil Moisture-Based
Irrigation Control Technologies
Table of Contents
Brian Koblenz, Irricloud	3
Peter Lackner, Toro Irrigation Division	5
Bernard Cardenas-Lailhacar, University of Florida	6
Celine Benoit, Metropolitan North Georgia Water Planning District
	7
Bob Beers, HydroPoint	9
Sean Steffensen, California Energy Commission	10
2	January 23, 2020

-------
EPA
Preliminary Comments on the WaterSense
Draft Specification for Soil Moisture-Based
Irrigation Control Technologies
WaterSense
The comments provided below are replicated in their entirety as submitted by each
stakeholder. The only changes made were of a typographical nature or for clarity, as
indicated by items in brackets. Any parenthetical items, emphasis through capitalization,
or grammar that is inconsistent with WaterSense style were provided by the
commenters.
Commenter: Brian Koblenz
Affiliation: Irricloud
Comment Date: November 25, 2019
Email Text:
I have both read your specification and participated in your webinar.
Most of my response is targeted to make sure that this process does not inhibit or
reduce innovative ideas that may improve water savings now and in the future. My
concern is that approving a specification like this IMPLIES that SMS [soil moisture
sensor] systems that do not meet this specification are likely inferior in terms of water
savings and, that should not be the implication.
FWIW [for what it's worth], I believe my company's products can meet all of your
specifications, however the lack of definition of "Irrigation Event" and what it means to
stop an "Irrigation Event" gives me some concern. (More on this later.)
You state systems that enable and disable irrigation events are "in-scope" yet you
exclude from your scope "On-demand SMS" which are precisely those systems that
enable and disable irrigation based on SMS data.
What I believe you are doing in actuality is ONLY including those systems that "disable"
pre-defined irrigation events and the absence of "disabling" the event allows the event to
move forward. In other words, the base controller "clock" enables the event and the
sensor can "disable" the event. In this scenario, SMS are given high thresholds above
which they must "disable" irrigation events. (I suspect most of the add-on SMS systems
rely on this scenario but they will struggle with the requirement to "notify" the base
controller that the SMS is not operating properly but I will leave it to those vendors to
make their case.)
One could imagine a different (better?) system where a clock has an "irrigation event"
program with a set of zones to run but no specified start time. Then the SMS could
interact with the clock to "enable" that program that would then run to completion. In this
scenario, the SMS are given a low threshold below which an irrigation event must be
"enabled". There are many things to recommend this style and it can work in the context
of "restricted watering days", but that restricted day framework may itself not lead to the
most efficient watering scenarios and may change over time.
One could also imagine yet another system where the "irrigation event" program is
based on a combination of clock information and SMS data where both starting and
stopping (enabling/disabling) the irrigation event is controlled by the SMS.
3
January 23, 2020

-------
EPA
Preliminary Comments on the WaterSense
Draft Specification for Soil Moisture-Based
Irrigation Control Technologies
WaterSense
All of the above scenarios are practical, useful, and efficient; supporting one to the
exclusion of the others is inappropriate.
What is an "irrigation event"? Most irrigation controllers today work with the notion of a
program and even within that term, different companies do things differently. A program
might:
a)	run 1 zone for some time
b)	run multiple zones concurrently for some time
c)	run a set of zones in sequence where each zone has a run time
d)	might mix b) and c)
e)	might "water and soak" any of a), b), c), d) so that the watering repeats some
number of times (usually within a day) at some interval
f)	let your imagination roam
My point is that I think you want to be tolerant of many different ways of skinning the cat
and your specification is insufficiently tolerant. In any case, I believe it is necessary to
clearly define irrigation event.
I don't want to set your process back too far, but maybe there is a way to be both simpler
and more tolerant.
You already have a test methodology for WBIC [weather-based irrigation
controllers]. How about creating your engineered boxes with a pre-defined moisture
content and place them in your test area where they are exposed to the weather and can
control an irrigation valve. Place the SMS in the box, connect it to the base controller
which can enable the valve and log the amount of watering that occurs. (Ensure the
base controller does not have access to the weather data or program it to be at some
very different location.) Keeping with the WBIC model, if, after a month (and assuming at
least 4 days of .1 inches of rain etc), the amount of watering is "good", then we have a
WaterSense approved system. You can still do the freeze test or any other stress you
want to create on the SMS but you are not limiting the way the base controller and SMS
interact.
I am sure the above could have been written more eloquently and I am happy to answer
any questions or discuss and clarify via email or phone.
-brian
brian.koblenz@irricloud.com
4
January 23, 2020

-------
EPA
Preliminary Comments on the WaterSense
Draft Specification for Soil Moisture-Based
Irrigation Control Technologies
WaterSense
Commenter: Peter Lackner
Affiliation: Toro Irrigation Division
Comment Date: December 13, 2019
Email Text:
We would like to request that a fourth moisture level, 100% (field capacity), be added to
the test protocol, at least for testing the Toro® Precision™ Soil Sensor, model PSS-KIT.
Per the instruction manual, this is required for the sensor's calibration procedure, as well
as for it to change states from allowing irrigation to blocking it while in operation.
Given that a) the testing facility will already have all required materials on-hand to make
this accommodation (water, soil, salt solution, and the tools required to calibrate to any
given percentage of moisture depletion), b) the test is only conducted once for any given
model of sensor, c) the manufacturer is paying for the test, and d) the dual-threshold
"checkbook" method of irrigating, which the Precision Soil Sensor uses, is an accepted
practice per the lA's [Irrigation Association's] handbook (see Irrigation 6th Edition, Ch. 13
"Irrigation Scheduling"), we feel this is a reasonable request.
In other words, the sensor does not operate incorrectly or inaccurately, it simply
operates differently from the other manufacturers that have been tested. As such,
creating the test such that it is designed to fail any existing product that already operates
in this manner might be considered unfair "restraint of trade" or undue burden on the
manufacturer to change its product.
Regards,
Peter Lackner
Product Marketing Manager
Toro Irrigation Division
5
January 23, 2020

-------
EPA
Preliminary Comments on the WaterSense
Draft Specification for Soil Moisture-Based
Irrigation Control Technologies
WaterSense
Commenter: Bernard Cardenas-Lailhacar
Affiliation: University of Florida
Comment Date: December 20, 2019
Email Text:
Hi All,
From the Public Meeting Summary comments, seems that some stakeholders are really
confused with the term "irrigation event" (there are two comments/questions on page 3
and two comments on page 4 regarding this concept). Sometimes seem that they are
referring to only one zone running and sometimes to all the zones that could potentially
run. I suggest changing the language to "irrigation cycle", and define this as "all the
irrigation zones that are programed to run sequentially after the first zone starts".
This definition encompass[es] the "on-demand" SMSs models that could start an
irrigation cycle at any time (not necessarily after a scheduled start time).
Regards,
Bernard Cardenas-Lailhacar | Research Associate |
Agricultural & Biological Engineering Dept. | University of Florida |
234 Frazier Rogers Hall | Gainesville, FL 32611 |
Q irrigation.ifas.ufl.edu |S : 352-392-1864 ext 234
6
January 23, 2020

-------
EPA
Preliminary Comments on the WaterSense
Draft Specification for Soil Moisture-Based
Irrigation Control Technologies
WaterSense
Commenter: Celine Benoit
Affiliation: Metropolitan North Georgia Water Planning District
Comment Date: January 7, 2020	
Email Text:
Please find attached the Metropolitan North Georgia Water Planning District's comments
for the WaterSense Draft Specifications for Soil Moisture-Based Irrigation Control
Technologies.
Feel free to reach out with any questions or comments.
Thank you for your consideration.
Water Efficiency Planner & GIS Analyst
Metropolitan North Georgia Water Planning District
International tower
229 Peachtree St NE, Suite 100
Atlanta, Georgia 30303
P: 470-378-1569
E: cbenoit@atlantareaional.org
Email Attachment
See page 8.
Hello,
Best,
(Odette Semit
7
January 23, 2020

-------
Metropolitan North Georgia Water Planning District
International Tower I 229 Peachtree St., NE I Suite 100 I Atlanta, GA 30303
WaterSense Draft Specification for Soil Moisture-Based Irrigation Control
Technologies; Comments
The Metropolitan North Georgia Water Planning District (The District) is committed to
continuing to support water efficiency not only on a regional scale within the Metro
Atlanta area, but nationwide. Our planning efforts and the implementation of water-
saving technology throughout our region has demonstrated positive impacts for our
communities. We continue to promote the use of WaterSense products in our planning
efforts and are pleased to see the inclusion of irrigation products such as soil moisture-
based controllers.
Please find below our comments regards the WaterSense Draft Specification for Soil
Moisture-Based Irrigation Control Technologies:
Under 3.0 Supplemental Capability Requirements
3.1 Is there a time frame established for how long content should be preserved within the
product giving a loss of power?
3.3	Are there further specification as to the equipment surrounding the notification
system to indicate when the system is not receiving sensor mechanism input? Does this
include notification via an app or on the physical system itself?
3.4	We support the requirement of ensuring the products capability of functioning with a
rainfall device, given that our Georgia Code (12-5-6) mandates irrigation systems
installed after 2005 using public water sources, be equipped with a rain sensor shut-off
switch.
3.5.2 and 3.5.3 Requiring day interval schedule is also supported by the District. Our
current Drought Response requires adopting an interval irrigation scheduling during
times of drought, with complete cessation at times of high drought response.
We look forward to the fmalization of these requirements as they develop, and are glad to
contribute in further discussions. Thank you for considering our comments.
Thank you,
Celine Benoit
Water Efficiency Planner & GIS Analyst
Metropolitan North Georgia Water Planning District
www. n o rt h g e o rg i a wa t e r. o rg

-------
EPA
Preliminary Comments on the WaterSense
Draft Specification for Soil Moisture-Based
Irrigation Control Technologies
WaterSense
Commenter: Bob Beers
Affiliation: HydroPoint
Comment Date: January 15, 2020
Email Text:
Good afternoon,
I hope you all had a great holiday season and new years [sic].
We attended the EPA session at the IA Show in [Las] Vegas regarding the moisture
sensor testing process and specification terms. You all have done a great job with
everything...we were very impressed!
One question we had as we left is related to verbiage around how a moisture sensor is
"connected to" or "communicates with" a smart irrigation controller. It sounded as though
the specification would require a moisture sensor to be directly wired to a smart irrigation
controller for it to qualify. With our Baseline Irrigation products we can support a
moisture sensor that is wired to a gateway device (we call it a Substation) which
wirelessly networks back to the smart irrigation controller. So, the moisture sensor still
delivers all of the real-time moisture data back to the controller which can act on the
moisture data with no measurable difference in reaction time versus a wired device.
We are hoping for some clarification around that scenario. It's the same moisture sensor
product and the same smart controller product...we just add in a wireless gateway so we
can manage a moisture sensor anywhere on the site.
Thanks!
Bob Beers
Product Manager, Irrigation Technology
o 208-639-8738 m 208-703-7141
HydroPoint 360° Smart Water Management
hvdropoint.com | baselinesvstems.com
9
January 23, 2020

-------
EPA
Preliminary Comments on the WaterSense
Draft Specification for Soil Moisture-Based
Irrigation Control Technologies
WaterSense
Commenter: Sean Steffensen
Affiliation: California Energy Commission
Comment Date: January 21, 2020	
Email Text:
The California Energy Commission provides comments in the enclosed letter.
Thanks,
SEAN STEFFENSEN, P.E. | MECHANICAL ENGINEER
CALIFORNIA ENERGY COMMISSION | EFFICIENCY DIVISION
1516 9™ ST, SACRAMENTO, CA 95814
(916) 651-2908 OFFICE |FAX (916) 654-4304
Email Attachment
See pages 11 through 15.
10
January 23, 2020

-------
CALIFORNIA
ENERGY COMMISSION
£
C*1IF3«\
resources
January 21, 2020
Ms. Stephanie Tanner
U.S. Environmental Protection Agency
Office of Water
WaterSense Program
1200 Pennsylvania Avenue, N.W.
Washington, DC 20460
Dear Ms. Tanner:
The California Energy Commission (CEC) appreciates the opportunity to
provide comments on the U.S. EPAs (EPA) proposed specification for Soil
Moisture-Based Irrigation Control Technologies (SMBICT). The CEC is the
primary energy policy and planning agency of the State of California. One of
the chief mandates of the CEC is to reduce the wasteful, uneconomic,
inefficient, and unnecessary consumption of energy and water in the state by
prescribing standards for minimum levels of operating efficiency for
appliances that consume a significant amount of energy or water on a
statewide basis. We recognize the importance of working closely with the
EPA to lead efficiency efforts that will incentivize energy and water efficient
technologies that will reduce the wasteful consumption of energy and water.
The CEC appreciates the EPA's efforts to establish a new voluntary WaterSense
specification for SMBICT especially as the state recovers from severe drought
conditions and continues to focus on ways to conserve its limited water supply.
The CEC is pleased that the EPA's specification proposes to address inefficient
irrigation scheduling - applying water when not needed. The SMBICT will
measure soil moisture content and prevent the wasteful application of water. The
specification has the potential to encourage consumers to choose products that
will automate irrigation and save what the EPA estimates to be hundreds of
billions of gallons of water across the country.
The CEC supports the EPA's proposed modifications to the American Society of
Agricultural and Biological Engineers (ASABE) X633 Testing Protocol for
Landscape Soil Moisture-Based Control Technologies to reduce the test time
while maintaining the repeatability and reliability of the test procedure.
energy.ca.gov
1516 9th Street, Sacramento, CA 95814

-------
Ms. Stephanie Tanner
January 21, 2020
Page 2
Specifically, the CEC supports the EPA's proposal to modify the test
procedure:
•	Soil moisture testing only in moderately coarse media and saline
water,
•	Freeze testing only in moderately coarse media and saline water at 40
percent water depletion,
•	Clarification to connect add-on and plug-in devices to a base controller
during testing as specified by the manufacturer.
The CEC encourages careful review of the calculation methods and
performance levels selected by the EPA. The CEC recommends some
modifications to ensure repeatability and clarity. The CEC provides this
information and recommended changes to the specification language in the
appendix to this letter.
The CEC appreciates the opportunity to comment on this draft specification. If
there are any questions about the attached comments, please contact Sean
Steffensen at (916) 651-2908 or at Sean.Steffensen@energy.ca.gov.
Sincerely,
DAVID HOCHSCHILD
Chair

-------
Appendix
Topic 1: Clarify the calculation of relative average deviation (RAD) as shown on slide
35 of webinar presentation and section 2.2.1.2 of the draft specification.
The draft specification says that the RAD will be averaged across all water depletion
levels. The WaterSense Draft Specification for Soil Moisture-Based Irrigation Control
Technologies (November 20, 2019) webinar slide 35 shows that the RAD should be
calculated at each water depletion level (20 percent, 40 percent, and 60 percent). The
webinar and the draft specification seem to be inconsistent.
The calculation of the average RAD across all water depletion levels could be made
clearer by adding Equation (3) to the specification.
The draft specification must also define the units of the performance criteria. The
webinar suggests the units are "sensor reading percent full scale."
Equation (3): RADavg=(RAD2o+RAD4o+RAD6o)/3
Where RAD20 is the relative avg deviation at 20 percent water depletion
RAD40 is the relative avg deviation at 40 percent water depletion
RAD60 is the relative avg deviation at 60 percent water depletion
Figure: Plot Showing Calculation of Relative Average Deviation





	y* 1..HO/X+ r.rrn
R' ¦ 0972
0
1—
]~10


1




AD = (10+0+10)/3 = 6.667
f"
«
RAD = 6.
667/90 = 7.4°/






Depletion Level
• Sensor 1 • Sensor 2 • Sensor i • Sensor Avg. 		Linear (Sensor Avg.)
Source: U.S. EPA WaterSense Webinar, November 20, 2019, Slide 35
Section 2.2.1.2 is included below for reference.
"2.2.1.2 The relative average deviation (RAD) of the readings at which the
replicate SMSs enable and disable irrigation, calculated in accordance with
Equations 1 and 2 below, when averaged across all water depletion level
readings, shall be less than or equal to 10 percent."

-------
Eqimtion (1)
Relative Avg Deviation =
Where: x is the mean
Avg Deviation
x
Equation (2)
Avg Deviation |= ^'=1^
Where: x is the mean
Xj is the observation
n is the number of observations
Topic 2: Clarify Equation 2 in section 2.2.1.2 to show that the calculation is at a single
water depletion level.
The test procedure requires three observations per each water depletion level.
Therefore "n" in equation 2 will always be three. Equation 2 could be made clearer by
replacing n with three in Equation 2.
Clarity also could be added by rewriting Equation 2 as a simple sum as shown below.
Suggested Equation (2) Average Deviation = [(x-x-i)+(x-X2)+(x-X3)]/3
Topic 3: The sensor readings and calculation methods in section 2.2.1.3 need to be
identified.
The draft specification provides this instruction to find the slope as a verification of the
device performance. The draft does not identify the readings or describe how the three
readings at each of the three water depletion levels are used to calculate the slope.
Section 2.2.1.3 "The absolute value of the slope of the line generated by plotting
irrigation enable readings for all three replicates across all three depletion levels
and the absolute value of the slope of the line generated by plotting irrigation
disable readings for all three replicates across all three depletion levels shall both
be greater than zero when rounded to two significant digits (i.e., > 0.01)."
The EPA must identify the irrigation enable and irrigation disable readings to remove
ambiguity. Are the readings a sensor value presented as a percentage of full scale? Is
the reading a resistance, a current or voltage? What units are used to record the
reading? If the reading varies among the soil moisture sensing technologies, then the
specification must define the differences.
The calculation result is sensitive to the units of the readings' measurement. The
performance criteria should be expressed in the desired units. If the readings are in
ohms then the criteria would be expressed as >0.01 ohms per percent water depletion.
Where: x is (xi+X2+X3)/3
xi is the first observation
X2 is the second observation
X3 is the third observation
2

-------
If units of measure vary among the soil moisture sensing technologies, then the
specification must define the units of the performance criteria for each technology.
The draft specification must define the vertical axis as the sensor reading and the
horizontal axis as the water depletion level so the slope can be calculated consistently.
The draft specification must define how the depletion level percentage is represented
when the calculation is performed. For example would "20" or "0.2" be used to represent
20 percent when the slope is calculated?
The draft specification does not provide a calculation method for the slope of the line.
Slide 37 of the webinar presentation shows a Microsoft Excel plot where the slope is
observed as the coefficient of the "x" value of the linear least squares fit of the data. The
draft specification must identify the linear least squares fit as the calculation method to
ensure consistency.
25.0
|20.0
a»
_3
re
> 15.0
¦o
—
-Q
ro
c.
iu
o 10.0
c
at
tn
5.0
0.0
0	20	40	60	80	100
Depletion Level
• Sensor 1 • Sensor 2 • Sensor 3 • Sensor Avg. 	Linear (Sensor Avg.)
Source: U.S. EPA WaterSense Webinar, November 20, 2019, Slide 37
Figure: Plot Showing Slope of Sensor Readings
«
1
Sensor Enabled
1


Slope
L


v[
A I
?.2542x | 26.778
1
"*• fl



R^a9548
't
~











3

-------