The FOCE prototype design consisted of a series of subsystem components: Frame, Directional Control Valve, Fluid mixing and Distribution, Control System hardware and software, and pH detection and feedback. Each subsystem is itself a novel prototype system for engineering design and testing evolution.
Prototype System Design.
All system components, and the final assembly, were tested in the MBARI 10m tank prior to field deployment.
The basic concept of a circular array of emitters used in FACE experiments was kept. The selected design consisted of two concentric rings, 0.5 m high, set 0.5 m apart. The inner ring (1 meter diameter) supported eight pH electrodes, mounted vertically. The outer ring (2 m diameter) supported the 40 fluid emitter hoses arranged in eight sectors and set vertically around the perimeter. The frame was a simple 6061 aluminum weldment with cross wire bracing.
Directional control valve.
Time and budget constraints forced a simple system for prototype testing. We used a simple mechanical weather vane for passive current detection. This turned a valve outer rotor made from PEEK polymer that opened and closed ports on the inner stator as it moved in response to flow, directing fluid to two adjacent sectors from a total of eight ports on the stator assembly. The vane was about 1600 cm2 area, and was of near zero buoyancy when deployed.
Fluid Mixing and Distribution.
For large volume and long term deployments delivery of the CO2 enriched sea water for pH control would have to be via a small bore pipe alongside the power/data cable. For the prototype short term experiments we used diluted HCl contained in eight 20 liter cubitainers mounted atop the frame. The selection of the valve port for emitting fluid was controlled by the central vane. The rate at which fluid was emitted was under active system control. Two peristaltic pumps were used; one delivered acid to a mixing chamber, the second drew in local sea water. A Hall effect turbine flow rate sensor was placed in-line between each pump and the mixing chamber, and the data from these sensors provided feed-back for control of the acid-seawater ratio that was emitted.
Eleven pH sensors were used on the FOCE frame: eight are deployed around the inner ring such that there tips are mid-way between the sea-floor and the top of the ring; three sensors are deployed at the center of the inner ring equally spaced between the sea-floor and the top of the ring. The information from the pH electrodes was fed into a central control unit. Only the electrodes supported in the center of the control volume were used for feedback and control in this prototype version; the electrodes on the outer ring were passively logged by a Seabird CTD, and data downloaded after the experiment. The controller presents data to the user via a graphical user interface (GUI) created in LabVIEW. The pH data were assimilated with a box car filter (first in first out) so as to smooth the data for analysis and control.
GUI Interface and Experimental Control.
The GUI allows the user to set the essential experimental boundaries such as the required ΔpH, and the initial sea water-acid mixing ratio. pH electrodes are then normalized to equal values to compensate for any calibration offsets occurring during deployment. The ΔpH is fit to a preset function that non-linearly adjusts the acid mixing rates by the control software as the observed ΔpH varies further from the desired value. The resultant value is used to inversely adjust the sea water pump rate, thus maintaining a constant overall fluid flow through the valve assembly. The model takes the inevitable latency of the system into account; typically the system overshoots initially, driving a lower than required pH, and then compensates by steadily bringing the system back to the desired value. As local velocities change the pumping rates change, and the sequence of events towards stabilization occurs.
pH electrode Calibration and Deployment.
Three electrodes were deployed in the center of the prototype ring for real time experimental control. Each electrode was calibrated in the laboratory before each deployment. From these calibrations a slope and offset was determined and applied via the GUI to each of the three electrodes. It has been our experience that upon deployment a calibration offset occurs with each pH sensor that arises from liquid junction effects in the reference electrodes due to the large temperature and pressure changes taking place. A simple in situ calibration adjustment corrects for this problem. Additionally, there are several options for alternate pH and pCO2 sensors etc. for incorporation into future systems as desired by the PIs adopting the technology and as experimental goals and needs evolve.