Mooring design and simulation for a hand-deployable nearshore real-time sensor package
Jeff Sevadjian, Matthias Lankhorst, Uwe Send
Scripps Institution of Oceanography
Introduction
The purpose of this work was to design a mooring system for the Mini-Mooring real-time sensor package developed by the Ocean Time-Series Group at the Scripps Institution of Oceanography. The overarching goal was to design a reliable mooring system for nearshore (≤ 30 m water depth) environments, capable of withstanding relatively large waves (3 - 5 m) and strong currents (0.70 m s-1), while being lightweight and manageable enough to be hand-deployed by no more than two people from a small boat.
Static and dynamic modeling were conducted using OrcaFlex softare (Orcina, Inc.). Dynamic simulations were carried out under a broad range of ocean wave and currents conditions to determine the maximum vertical and horizontal tensions along the mooring line under various mooring configurations and under conditions typical to shallow coastal systems (kelp forests, coral reefs, estuaries, etc.). Tensions at the anchoring point of the mooring were compared to estimates of holding power for commonly used, commercially available anchors, to estimate the minimum-sized anchor required for a given field site based on the expected oceanographic conditions.
See also: Real-time data from a pilot deployment in Puttalam Lagoon, Sri Lanka
Mooring Design
The Mini-Mooring consists of a small surface float that houses controller electronics and satellite and inductive communications modems boards and GPS, and an inductive mooring wire to which a range of commercially available inductive oceanographic sensors is coupled (Fig. 1).
Subsurface sensor(s) are located between the sea floor and a subsurface float which keeps the lower section of the mooring line vertical. Above this, three small floats are attached to the mooring line to form a smooth bend radius, helping to keep the wire from becoming entangled. Diver observations during a preliminary field deployment in La Jolla, California confirmed this setup was effective at avoiding entanglement. The upper 2 m of the mooring line is kept vertical by two small weights attached to the mooring line.
Figure 1. General mooring design as modeled in OrcaFlex
Three separate configurations were investigated for the bottom portion of the mooring:
1) Inductive wire anchored directly to the sea floor;
2) Inductive wire connected to a length of chain, anchored to the sea floor;
3) As in (2), with a large clump weight added at the union of the inductive wire and chain
Results are presented below for each of these anchor configurations.
Simulations were run for a 10-m water column with waves up to 3 m, illustrative of a dynamic coastal reef environment, and a 30-m water column with waves up to 5 m, illustrative of a highly exposed, energetic outer reef.
Hardware Components
All controller electronics and batteries for the Mini-Mooring are housed within the surface float. During preliminary deployments in La Jolla, California, and Puttalam Lagoon, Sri Lanka, a 0.38 m diameter, 12.9 kg (dry weight including electronics and batteries), +16.6 kg buoyancy sphere from the Scripps Instrument Development Group (IDG) Argos program was used. Under the relatively benign oceanographic conditions of these field sites, this provided enough buoyancy to keep the buoy afloat without pulling the anchor. However, model simulations using this size float indicated the mooring would be knocked down and the float submerged under currents > 0.5 m s-1. While the surface float is water-tight, it would not be able to communicate with shore when submerged. Thus, for these simulations a larger float, modeled after one also available from Scripps IDG, was used. The larger float is a 0.48 m diameter sphere weighing 20.2 kg (dry weight including electronics and batteries) with +39.2 kg buoyancy.
The surface float was modeled as a 6 degrees-of-freedom Spar Buoy. For the purposes of the model the spherical float was represented as a series of 5 short cylinders stacked on top of one another with tapering diameters. The dimensions of the cylinders were defined such that the total volume was equal to that of the actual 0.48 m diameter sphere.
The mooring wire is 0.25 in (0.0064 m) outer diameter Nilspin jacketed wire rope. The wire is not only an integral structural component of the mooring but also the means by which the surface controller and subsurface sensors communicate. The mooring wire is 0.11 kg m-1 mass per unit length. The bending stiffness of this wire was not known, and in the model was set by manually adjusting this quantity until the layout of the mooring components in the static solution (e.g., Fig. 1) resembled that of in-situ diver observations; the value used for these simulations was 0.3 N m-2.
For the main subsurface float, a 12 in (0.30 m) diameter, +23.8 lb (+10.8 kg) buoyancy trawl float from Seattle Marine & Fishing Supply Co. (SMFS) was used. Above this, three, +1 lb (+0.5 kg) buoyancy gillnet floats from SMFS were attached to the inductive mooring wire at 0.30 m intervals to form the inverse catenary.
The mooring chain used in these simulations was 0.50 in (0.013 m) diameter Spectrum 3 carbon steel studless chain from The Crosby Group. A smaller, 0.25 in (0.006 m) diameter chain was initially considered, given the benefit of decreased weight for manual (by-hand) deployments, but this was rejected in favor of the larger 0.5 in chain which would be more structurally resilient to corrosion and decay over longer deployments. This 0.50 in chain has been used extensively by the Ocean Time-Series group for both open-ocean and nearshore moorings, with typical deployments lasting 1-2 years. This chain has a mass per unit length of 3.77 kg m-1.
To keep the overall weight of the mooring low, the anchor must be very efficient—e.g., capable of withstanding a high degree of tension before breaking out from the sea floor, while being relatively lightweight itself. For these moorings, with relatively low buoyancy and in shallow (≤ 30 m) water, a standard boat anchor is prefered over the large 'clump'-style anchors, such as railroad wheels, commonly used for larger moorings. By digging into the sea floor when dragged by winds or currents, these anchors are capable of holding large yachts and buoys many times their own weight. We compared modeled horizontal and vertical tensions at the mooring's anchoring point to estimates of holding power given by anchor manufacturers (e.g., Danforth) and from independent tests (Smith, accessed 2018; Knox, accessed 2018).
In addition, for the 30-m water column, exposed-site scenario, a large, iron clump weight was attached to the mooring at the union of the inductive wire and chain to dampen the load on the anchor. Initially, a 20 kg (dry weight) clump was used, but simulations of surface waves > 3 m suggested these waves would pick up the clump and then drop it rather violenty with each passing wave, causing snap loading. Incrementally larger iron masses were modeled until the mass was no longer moved by waves up to 5 m at 20 s, and currents up to 0.7 m s-1. This required mass was 120 kg (dry weight), or a buoyancy of -104.4 kg.
A SeaBird Electronics SBE-37IM was included in the model and given realistic values for mass, volume and height.
Numerical simulations
Simulations were run separately given a single wave train input of a particular height (1, 2, 3, or 5 m) and period (T = 10, 12, 15, or 20 s), as well as for different current speeds (0.1 to 0.7 m s-1, uniform throughout the water column), for 120-s runs. Time-series plots confirmed that tensions converged reasonably over this time frame. For each simulation, the maximum tensions at the anchoring point in the x- and z- (global reference frame) directions were recorded. In some simulations outliers associated with model spin-up were apparent; to avoid this bias only values after 3 full wave periods were considered. Other rapid spikes in tension apparently caused by a numerical artifact related to the mooring line discretization length were first smoothed with a sliding median of window size T/64 to find the index of the maximum, then the actual value at that index was recorded.
Results — 10-m water column
Results are presented below for each of the three anchor configurations. Time-series plots and animations for the dynamic solution of each simulated environmental condition can be accessed via the tables below.
With the inductive wire anchored directly to the sea floor, the modeled tensions at the anchoring point were below 30 kg for modest (1 - 2 m) wave heights (Fig. 2). For the 3-m wave, due to the relatively short scope of the inductive wire (17 m total length in 10 m water), the line became taut during the passing of each wave crest, causing both the vertical and horizontal tensions to increase substantially across all wave periods (see links to time-series plots and animations in the below table). Additional scope could be added, but at the risk of entanglement and abrasion of the subsurface sensor.
Tension in both x and z increased exponentially as a function of current speed, with the x component generally about twice that of z, to a maximum of 107 kg at 0.7 m s-1. This is well within the holding power of even a very lightweight Danforth anchor; for example, the holding power of a 35 lb (16 kg) Hi-Tensile Danforth anchor is quoted at 3800 lb (1724 kg). Independent tests of similar anchors estimated the 'ultimate holding capacity,' or maximum force an anchor will hold without moving in the seabed, of a 18.3 kg Spade anchor at 420 kg.
Of note, the +10.8 kg sub-surface float was knocked down by over 45° even under modest currents of 0.3 m s-1 (see links to static solution images in the below table); a larger float would keep the line more vertical but would require a larger anchor.
Summary
Figure 2. Maximum horizontal (lines) and vertical (markers) tension at the anchoring point under various wave and currents conditions for Mooring Configuration 1. Colors correspond to wave heights (see legend) plotted as functions of wave period; black corresponds to currents.
| 10 s | 12 s | 15 s | 20 s | |
|---|---|---|---|---|
| 1 m | | | | | | | | |
| 2 m | | | | | | | | |
| 3 m | | | | | | | | |
| 0.1 m s-1 | 0.3 m s-1 | 0.5 m s-1 | 0.7 m s-1 | |
|---|---|---|---|---|
The 8-m chain leader provided additional slack to the overall mooring length and dampened forces at the anchoring point. Tensions decreased by 50% (horizontal) and 92 - 97% (vertical) in 3-m waves compared to Configuration 1. The density of the chain kept it mostly resting along the sea floor under quiescent conditions, unfurling as the wave crests passed and only actively lifting up the links closest to the anchoring point at the very peak of a 3-m wave.
The horizontal tension under currents was within 1 kg to that of Configuration 1 for all current speeds, but the vertical tension under currents was 78 - 98% lower.
As stated in the Hardware Components section [above], with a relatively smaller clump weight, waves would pick up the clump and then drop it violenty, causing snap loading. To avoid this, a larger mass was used such that the clump weight was no longer moved by waves. However, this had the effect of removing compliance from the lower half of the mooring, and under larger waves the full length of the mooring line became taut after each passing wave crest, jerking the surface buoy [see videos linked below].
In the time-series plots linked via the table below, tensions are plotted both at the anchoring point, and at the location of the clump weight.
Results — 30-m water column
For the 30-m water column, exposed-site scenario, only Configuration 3 was tested.
As in the 10-m water column simulations for Configuration 3, fixing the clump weight to the sea floor caused the mooring line to become taut after each passing wave crest, resulting in high tensions on the buoy [see videos linked below.]
In the time-series plots linked via the table below, tensions are plotted both at the anchoring point, and at the location of the clump weight.
