Solar Based Automatic Irrigation Using Soil Moisture Sensor

Replies

• 

  Ramani Aswath

  In my opinion sun tracking or concentrating are contra indicated for the intended application. For one thing the cost goes up five to ten times for robust tracking or concentrating systems. For another they are high on maintenance, which is a problem in remote areas.

  For any reasonable irrigation system one needs a minimum of twenty square metres of solar PV panels. Difficult to use tracking or concentration.

  Since it is an automatic system irrigation can take place at any time. Between 10 am and 2 pm the efficiency of insolation is between 85% to 100%. Even at 9 am and 3 am it is 70%. The implication is that tracking is not that important. One can choose to irrigate at maximum insolation period (10 am - 2 pm).

  Maybe you can concentrate ( pun slipped in) on the software. You can divide the area into sectors with individual sensors and selectively deliver the water. You may have to allow for the time delay between delivering the water and the sensor to react to it after the water diffuses to it. Maybe some intermittent watering would be an optimum solution.

• 

  Alice Zan

  Good day sir #-Link-Snipped-#,

  Thank you for spending your time to reply my post.

  Sir, 

  To summarize your suggestion, tracking and concentrator is not required as both could be very pricey.

  On the other hand, the irrigation system should operate at certain times where the solar energy is optimum.

  However, I didn’t understood completely about the last paragraph (except the pun). Can you please elaborate about the time delay and the software?

  Again,

  Thank you, sir.

• 

  Ramani Aswath

  #-Link-Snipped-# , we have to look at the dynamics of the process. Different plants have different root patterns. Not all need continuous watering. Sugarcane for example is traditionally watered once a fortnight, while rice is cultivated using wasteful flood irrigation. It has been shown that 1 kg rice represents 2000 litres of water! An adult rice eater can be using a whopping 12 m^3 of water a month.

   Studies have shown that sprinkler systems for rice irrigation can boost production by 50% while cutting down water consumption by 75%.

  Irrigation needs vary depending on many local variables and season. Also a large field may have different water requirements at different sections. A single sensor will be a poor way of decision making.

  What is needed is an expert system that uses inputs from sensors, seasonal needs, local field specific information to ensure optimum deployment of available water. Adding internet of things to it can generate invaluable data for enhancing AI for continued improvement.

  The irrigation hardware (solar panels, pump and plumbing) is fairly straightforward. It’s the expert system that will make the major impact.

• 

  Ramani Aswath

  I forgot about the time delay bit.

  Soils have variable water diffusion rates. Sandy soils drain fast while clayey soil is slow to percolate. When a sensor gives a signal for water need it may be better to sprinkle water for a short period and wait a while monitoring the sensor. Short timed bursts may be better than continued watering till the sensor reads okay.

  As said earlier a well conceived expert system can handle all such issues.

  Actually the project can well be designing this system based on discussions with local agricultural institutions should pay dividends.

  #-Link-Snipped-#

• 

  Ramani Aswath

  Malaysian Agricultural Research and Development Institute | Land Portal