ckunte.net

Flooding

For steel jackets designed to be floated and upended, a common engineering practice is the consideration of a suitable percentile of reserve buoyancy. Typically, it is 10% in intact condition, and 5% in damaged (accidental) condition. Here are some stats:

Reserve buoyancy Jacket weight v. Intact reserve buoyancy

Improvements in the industry with regards to robust leg diaphragm design, in which they are tested for twice the design pressure, combined with adequate procedures in-place — viz., inspection, pressure checks, etc., prior to loadout — provide a degree of confidence that such an accidental damage can indeed be managed, and that it need not be designed for damage entirely. In support of such a premise, this section makes the case that the time taken to flood a damaged jacket leg, a commonly ignored factor, has the potential to render such a risk to as low as reasonably practicable.

In order to apply this ALARP premise, the following considerations need to be satisfied:

  1. A rigorous inspection and pressure tests of legs, flood valves, and diaphragms are carried out prior to loadout.
  2. Adequate care is taken to avoid any damage to diaphragms and flooding valves during tow (sea transport).
  3. Any damage that may yet occur, despite the above, would not be visible to the naked-eye when inspected, otherwise it would still be possible to consider mitigation prior to lowering the jacket in water. (Alternatively, it may be the case of flood valves leaking. It is further possible to maintain a passive pressure to keep water ingress in check, which is all possible only if the leak is small, and does not overcome the maintainable pressure and the hardware.)

Under such circumstances, the time taken (in seconds) to flood a damaged jacket leg completely may be calculated, e.g., as follows:

\begin{aligned} t = \frac{V}{Q} \end{aligned}

where, V is the total inner volume of a jacket leg (in m3), and Q is the rate of water ingress in to it (in m3/s), which is given by:

\begin{aligned} Q = v \cdot A \end{aligned}

where, A is the area of (diaphragm) rupture, and v is the velocity of water ingress (in m/s), which is given by:

\begin{aligned} v = \sqrt{2gh} \end{aligned}

where, g is acceleration due to gravity (typically 9.81m/s2), and h is the hydrostatic head (in m), which is calculated as follows:

\begin{aligned} h = \frac{p}{\rho} \end{aligned}

where, $\rho$ is the unit weight of seawater, and p is the hydrostatic pressure per leg, which is given by:

\begin{aligned} p = \frac{(W_w - r_b)}{a} \end{aligned}

where, rb is the reserve buoyancy of jacket per leg, a is the projected one leg area to free water surface, and Ww is the weight of water inside a fully flooded leg, which is given by:

\begin{aligned} W_w = V \cdot \rho \end{aligned}

Example problem

The following is an example of a 4 legged jacket, with a reserve buoyancy of 218t, with an inner volume of one of its typical legs 92.34m3, and considering one of its submerged legs with a damage of 25mm equivalent diameter):

Units:
ton = 1000 kgf        # Unit conversion
  g = 9.81 m/s^2      # Acceleration due to gravity
rho = 10051.82 N/m^3  # Unit weight of sea water (=1025 kgf/m^3)

Total reserve buoyancy of jacket (average of intact and damaged conditions):

rbt = 218 ton

Number of jacket legs:

n = 4

Projected one leg area to free water surface:

a = 130.78m^2

Inner volume of jacket leg:

V = 92.34m^3

Weight of water inside the fully flooded leg:

Ww = (V * rho) in N => 928,185.059 N

Reserve buoyancy of jacket per leg:

rb = rbt / n in N => 534,462.425 N

Equivalent rupture hole diameter:

d = 25mm

Area of rupture:

A = (pi / 4) * d^2 => 490.874 mm^2

Hydrostatic pressure per leg:

p = (Ww - rb) / a => 3,010.572 N/m^2

Hydrostatic head:

h = p / rho => 0.3 m

Velocity of water ingress:

v = sqrt(2 * g * h) => 2.424 m/s

Rate of water ingress into the jacket leg:

Q = (v * A) in m^3/s => 0.001 m^3/s

Time to flood jacket leg completely:

t = V / Q in hours => 21.556 hours

Consider the same problem, with a diaphragm rupture area of 100mm diameter equivalent instead, and we get:

Rate of water ingress into the jacket leg (m^3/s): 0.019
Time to flood (hours): 1.3472

Notice the dramatic reduction in time taken to flood a leg, the logical reasoning here being that a jacket with an equivalent diameter of 100mm of damage should never be put in water, because it would be visible simply to the naked eye, and would allow the installation team to develop a fix prior to lowering the jacket in water.

With flooding being a transient condition, if it can be further assumed that the initial diameter tends to grow larger in time to a final diameter (equivalent area) by the time hydrostatic head is equalized, then the average time taken to flood would be about 11hours.

Now, if the duration of flooding the entire damaged jacket leg is long enough to complete flotation and upend installation sequence, like in this example above, then it can be reasoned that the damage may not necessarily need to be mitigated since it is deemed not safety critical — in terms of time.

Reason for this is obvious: eliminating or minimizing the need for buoyancy tanks, which are both painful and expensive to install and later remove post the jacket upend.