Estimate of Weather Window for marine operations, based on significant wave height

Estimate of Weather Window for marine operations, based on significant wave height#

The objective is to provide the probability of occurrence and expected mean number of events of weather windows corresponding to situations when \(Hs\) remains below an access threshold (\(Ha\)) for a given duration \(T\). Associated operational parameters are the access time and waiting time. Because of the seasonal variability of the wave climate, Weather Windows parameters are estimated over monthly periods. The implemented method is based on results presented in:

  • EQUIMAR Deliverable D7.4.1 «Procedures for Estimating Site Accessibility and Appraisal of Implications of Site Accessibility » (2010), T. Stallard, University of Manchester, UKJ-F. Dhedin, Sylvain Saviot and Carlos Noguera Électricité de France, France)

  • Walker RT, Johanning L, Parkinson R. (2011) Weather Windows for Device Deployment at UK test Site: Availability and Cost Implications, European Wave and Tidal Energy Conference, Southampton, EWTEC2011.

Created on 24/03/2021

@author: Christophe Maisondieu

class resourcecode.weatherwindow.WeatherWindowResult(weibull_distribution_result: WeibullDistributionResult, tau: ndarray, PT: ndarray, number_events: ndarray, number_access_hours: ndarray, number_waiting_hours: ndarray)#

The resulting estimation

Parameters:

  • weibull_distribution_result (WeibullDistributionResult) – Weibull adjustment parameters

  • tau (np.ndarray) – Persistence: Mean duration of periods for which Hs<Ha (Hours)

  • PT (np.ndarray) – The probability of occurence of a weather window corresponding to sea-state having hs < hs_access_threshold

  • number_events (np.ndarray) – number of events

  • number_access_hours (np.ndarray) – number access hours

  • number_waiting_hours (np.ndarray) – number waiting hours

class resourcecode.weatherwindow.WeibullDistributionResult(Ha: ndarray, x0: float, b: float, k: float, _MCFrHS: ndarray, _X: ndarray, _Y: ndarray, _residual: float)#

The resulting fitted parameters for a Weibull distribution

Parameters:

  • Ha (np.ndarray) – the values for which the distribution has been fitted

  • x0 (float) – the parameter of the Weibull distribution

  • b (float) – the parameter of the Weibull distribution

  • k (float) – the parameter of the Weibull distribution

P#

P is the probability of exceedance P(Hs > Ha)

Type:

np.ndarray

Ha#

the values for which the distribution has been fitted

Type:

np.ndarray

x0#

the parameter of the Weibull distribution

Type:

float

b#

the parameter of the Weibull distribution

Type:

float

k#

the parameter of the Weibull distribution

Type:

float

resourcecode.weatherwindow.compute_weather_windows(hs: Series, month: int, hs_access_threshold: ndarray | None = None)#

Identification of weather windows

Montly statistics NMI Method

Based on EQUIMAR Deliverable D7.4.1 Procedures for Estimating Site Accessibility and Appraisal of Implications of Site Accessibility (T. Stallard,University of Manchester, UKJ-F. Dhedin, Sylvain Saviot and Carlos NogueraElectricité de France, France) August 2010

and

Walker RT, Johanning L, Parkinson R. (2011) Weather Windows for Device Deployment at UK test Site: Availability and Cost Implications, European Wave and Tidal Energy Conference, Southampton, EWTEC2011.

Parameters:

  • hs (a pandas Series given the Significant Wave Height (m)) – with a datetime index.

  • month (int, the month number for which the weather window should be) – computed.

  • hs_access_threshold (an optional numpy array given the significant wave) – height operational access threshold

Returns:

results

Return type:

a WeatherWindowResult

resourcecode.weatherwindow.fit_weibull_distribution(hs: Series) → WeibullDistributionResult#

Fit a Weibull distribution on Hs.

The probability of excedence \(P(Hs > Ha)\) follows a Weibull distribution given by three parameters:

\(P(Hs > Ha) = exp(-((Ha - x0)/b)^k)\)

Parameters:

hs (a pandas Series) – giving the Significant Wave Height (m) with a datetime index.

Returns:

weibull_distribution_result

Return type:

WeibullDistributionResult