Tensiometer-based irrigation in highly-amended media.

Prof. Heiner Lieth

Environmental Horticulture, UCDavis, Davis, CA 95616

May 7, 1997

(Send comments/questions/suggestions to jhlieth@ucdavis.edu; for best printout set browser to 12pt Arial font, 0.7inch margins)

The purpose of this document is to identify and discuss technical elements regarding tensiometer-based irrigation in a setting where the sensors are able to register changes in moisture tension rapidly (i.e. in seconds, rather than minutes). Conventional field tensiometers installed in field soil are usually too slow to operate this way.

The information below is aimed at greenhouse production of plants in pots or in the ground, where the rooting medium is either a very-porous, artificial medium, or soil that has been highly amended to have higher water-holding capacity than is common in field soils.

Tensiometers deal with moisture tension or matric potential of some porous medium (such as soil or potting mix). In scientific jargon, "martic potential" is just one concept of a wide range of terms dealing with water pressure in soils and plants. It is in "pressure units" - positive values indicate pressure, negative values indicate suction or tension. Thus the matric potential of soils or potting mixes generally will be negative. The term moisture tension represents the degree of such suction (and is a positive number). For example, if the matric potential of a soil is -10 kPa (kPa=kilopascal), then the moisture tension is 10 kPa. This is pretty simple but can lead to confusion when scientists and growers get together and start talking about high or low tensions and potentials.

In this discussion I will use only the term "tension" (rather than matric potential). Thus all the numbers will be positive. The higher the number, the higher the tension, the dryer the condition. The lower the number, the lower the tension, the wetter the conditions.

Moisture tension in pots

Before one can fully understand the issues and the optimal strategy for irrigation based on SMT (soil moisture tension) it is necessary to understand how tension is related to water content and how this varies with depth of the root-zone. The diagram below (Fig 1) shows the relationship between mositure tension and moisture content for the medium UC mix. The general shape for most potting media is similar, but they do vary enough to affect final recommendations for how to irrigate.

Note that this curve illustrates some general points about irrigating plants. The wet condition (i.e after a thorough irrigation) is for tensions near zero. As water is removed from the pot by the plant (or by evaporation) the status progresses along the curve to higher tensions and lower water contents. Note that this particular mix can have a water content of 75%. So if we have a one-liter (1000 ml) container filled with this medium, then this can hold 750 ml of water. But note that a significant portion of this (220 ml) is not available to plants, so that only 530 ml is available water.

Another thing to note is that much of this available water is exhausted by the time the dry-down has reached 7 kPa. At this point, extraction of a little more water sends the tension over 10 kPa and beyond. At tensions over 10 kPa a plant which is accustomed to fairly moist conditions will start showing signs of wilting. Unless one starts irrigation fairly soon, the plant will be exposed to conditions which are damaging. Thus it is generally wise to irrigate when the tension is around 5 kPa; after that the urgency increases radically with increasing tensions.

Note that the diagram shows which ranges of tension are best for the plant (1 to 5 kPa) and that tensions above 10 kPa are dangerous. The diagram also indicates how a human might percieve soils at various moisture tensions. Note that at 10 kPa one will still be able to sense moisture in the medium.

The sensor:

A tensiometer is a device that can measure moisture tension in a porous medium. The porosity of the ceramic tip dictates how fast water moves in and out of the instrument. Basically, if the instrument will encounter very dry conditions (high tension), then the ceramic needs to be very fine, making it very slow. If the tensiometer will be not encounter dry conditions, then it can be made of more-porous ceramic which will react much faster.

The traditional tensiometer comes equipped with a dial-gage. The system discussed here assumes that this gage has been replaced (or supplemented) with a transducer which translates the suction inside the tensiometer into an electrical signal which can be calibrated to represent the tension. With this conversion, the tensiometer can be attached to automated control systems.

Calibration is not a simple matter and involves aspects which probably must remain under the domain of the operator. (See Calibration below)

Tensiometers in pots

The following diagram shows a pot containing potting medium and a tensiometer at various moisture conditions.

The basic concept is that the weight of the water in the tensiometer pulls on the gage, as does the water in the medium. It is not the quantity of water that affects the measurement, but the length of the "water column". In other wrods, the "pull" on any point in the system is related to the distance that that point is from the bottom of the pot. For each centimeter above the bottom, the tension rises 0.1 kPa. Thus each 10 cm of water column (ie. depth) corresponds to 1 kPa of tension. (This is a case where using metric units makes things much easier).

In the example (Fig 3) the medium in the pot is 13 cm deep; the water column in the tensiometer is 7 cm long and the top of the ceramic tip of the tensiometer is 9 cm above the bottom of the pot. If the pot is at saturation (ie. you cannot get anymore water to be in the pot without having water drip out the holes in the bottom), then the tension at the bottom of the pot is 0 kPa; at the ceramic tip (9 cm higher up) it is 0.9 kPa; at the meniscus in the tensiometer it will be 1.6 kPa (16 cm above bottom). Thus, the lowest reading on this tensiometer is 1.6 kPa. I.e. this instrument cannot read zero in this setting.

Similarly, for any reading on the gage, one would subtract 0.7 kPa to determine the tension at the ceramic tip, or 1.6 kPa for the tension at the bottom. For example a reading of 5.4 kPa on the gage means that the bottom of the pot is at 3.8 kPa. Another example: if you want to irrigate when the ceramic tip is at 5 kPa, then you will do this when the gage reads 5.7 kPa.

SMT-based irrigation

While a lot of the information here is useful for either manual or automated irrigation, the objective here is to provide information on using tensiometers as sensors in automated irrigation, and to use SMT (soil moisture tension) to make irrigation decisions.

There are lots of different types of automated irrigation system. One fundamental assumption about this technology is that the system is capable of distributing water uniformly. If this is not the case then there will always be problems (regardless of whether the system is automated or not).

But even if the system is not uniform, one still has to make the decision as to when to irrigate and when to stop irrigation. Regardless of the system, tensiometers can always be helpful in this decision-making process.

System operation:

The basic operation is to have one sensor (tensiometer) to coincide with each irrigation valve. This sensor needs to be in a representative location in the crop being irrigated. The irrigation system should be uniform.

To make it work one needs two set-points:

  1. The high-tension set-point, which represents the level of dryness at which irrigation is warranted.
  2. A low-tension set-point represents the set-point at which an ongoing irrigation will stop.

These set-points are used as follows.

Typically plants will be extracting water from the root zone. As this occurs (usually over hours and days) the tension in the root-zone will gradually rise. Once the high-tension set-point is reached, the irrigation is initiated (in large scale operations this will mean that this particular irrigation circuit is scheduled for irrigation by placing it in an irrigation queue).

Once the irrigation is in progress, the tensions will drop (usually over seconds or minutes) as water is applied. The rate of water application should be slow enough so that the tensiometer can follow the change in moisture condition. When the low-tension set-point is reached, the irrigation is stopped. There is always some lag in this system so there will be some over-shoot.

If the high-tension set-point is such that the moisture content of the medium is relatively dry, then as new water is applied it may not move very fast laterally (sideways). This situation should be avoided since it will result in dry sections in the root-zone.

Other control parameters:

Automated control schemes should involve the following additional parameters:

Note that it is possible for an operator to set things up so that the high tension set-point is relatively high, the low-tension set-point is near saturation, and the max on-time is short. This can then result in situations where the tension does not drop down to the low-tension set-point. There are reasons why an operator might want to do this, but if it is done inadvertently, then it could lead to problems (salt build-up, part of crop drying out,...). It may be wise do build some sort of warning system into control software to track this sort of behavior and warn the operator of the potential danger.

Another relatively complicated pattern that should be warned about is if irrigation does not result in the expected behavior of the tension signal. Such poor behavior could mean some defect in the electronics related to the transducer or even in the control computer housing.


There are two aspects to calibration when using tensiometers. The first is to assure that the instruments measures tensions correctly, the second involves how the instrument is used. The first should be done by the firm that makes the instrument and the control equipment; the second is the responsibility of the grower. Another way to say this is: the vendor is responsible for assuring that the instruments behaves as illustrated in the diagram above, and the grower is responsible for using it correctly (i.e. identifying the proper reading to go with the tension set-point recommendations (provided below).

Calibration is accomplished by attaching a flexible clear tube to the lower portion of the tensiometer. This tube and the tensiometer are filled with clean water (there should be no bubbles). It is possible to have water columns of various heights; the difference in height between the meniscus in the tensiometer and the tube represents this water column. This length of water column is converted to tension as indicated above (10 cm = 1 kPa).


Extensive research has shown that keeping plants in a 1 kPa to 5 kPa range is optimal. Thus the high-tension set-point recommendation is 5 kPa; the low-tension set-point is 1 kPa. However, this information can be confusing if you consider that the height of the pot and length of tensiometer both affect tension readings. So what does this mean?

Here is the bottom line: whatever tensions are recommended, assume that they are "above saturation" and pertain to a location in the center of the root zone (usually at the porous tip). So the above recommendation is to have the root zone 5 kPa higher than what it is at saturation before watering, and to stop the irrigation when this point is 1 kPa above saturation.

The way you make this work is as follows. You insert the calibrated, water-filled instrument into the root medium in the pot. Now water thoroughly, making sure that all the medium is wetted completely. This may mean applying water serval times at half-hour intervals. The last time make sure the pot is sitting level, and wait until water stops dripping from the pot. At this point take a reading. This is your "saturation reading". Your high-tension set-point would be this value plus 5 kPa. This process of identifying the "saturation reading" is the grower calibration.

As an example consider the system diagrammed in Fig 3 (above). Here the crop would be growing in containers filled with a potting mix (e.g. UC Mix). The tip of the tensiometer is 9 cm above the bottom of the pot and the tensiometer tube is 7 cm long. As mentioned earlier, at saturation this container reads 1.6 kPa. Thus the high-tension set-point for this instrument in this situation would have to be 6.6 kPa (the recommended 5 kPa plus the "saturation reading of 1.6 kPa). The low-tension setpoint would be 2.6 kPa.

For more information or feedback contact jhlieth@ucdavis.edu