Vessel Emissions API Use Case¶

Run this example in Colab¶

EUA analysis¶

In this notebook, we will show a use case of estimating the number of EUAs needed for a fleet for the current year.¶

We will choose our fleet as a list of vessels, get their historical voyage emissions using Signal Ocean's Vessel Emissions APIn and use linear extrapolation to estimate the needed EUAs at the end of the year.¶

Setup¶

Install the Signal Ocean SDK:

pip install signal-ocean

Set your subscription key acquired here: https://apis.signalocean.com/profile

In [ ]:

pip install signal-ocean

pip install signal-ocean

Initialize Vessel Emissions API¶

In [1]:

signal_ocean_api_key = '' #replace with your subscription key

from signal_ocean import Connection
from signal_ocean.vessel_emissions import VesselEmissionsAPI

connection = Connection(signal_ocean_api_key)
emissions_api = VesselEmissionsAPI(connection)

signal_ocean_api_key = '' #replace with your subscription key from signal_ocean import Connection from signal_ocean.vessel_emissions import VesselEmissionsAPI connection = Connection(signal_ocean_api_key) emissions_api = VesselEmissionsAPI(connection)

Import helpful modules¶

In [2]:

from datetime import date, timedelta
from tqdm import tqdm
import pandas as pd
import datetime as dt
import numpy as np

from datetime import date, timedelta from tqdm import tqdm import pandas as pd import datetime as dt import numpy as np

Input Data¶

Choose a fleet as list of IMOs.¶

In [3]:

vessel_imos = [ 9299111, 9314193, 9411197, 9425526, 9440473, 9459096, 9461659, 9462354, 9486922, 9486934, 9487469, 9487471, 9487483, 9505819, 9524982, 9543536, 9580405,
9580417, 9688336, 9688348, 9723007, 9726619, 9767340, 9773947, 9780251, 9783992, 9833723 ]

vessel_imos = [ 9299111, 9314193, 9411197, 9425526, 9440473, 9459096, 9461659, 9462354, 9486922, 9486934, 9487469, 9487471, 9487483, 9505819, 9524982, 9543536, 9580405, 9580417, 9688336, 9688348, 9723007, 9726619, 9767340, 9773947, 9780251, 9783992, 9833723 ]

Use the emissions api to get historical emissions data for the list of selected vessels, including EU emissions. We will store the data as pandas dataframe.¶

In [4]:

emissions_list = []

for imo in tqdm(vessel_imos):
    vessel_emissions = emissions_api.get_emissions_by_imo(imo, include_eu_emissions=True)
    vessel_emissions = [emissions.to_dict() for emissions in vessel_emissions]
    emissions_list.extend(vessel_emissions)

emissions_df = pd.json_normalize(emissions_list, sep='')
emissions_df.describe()

emissions_list = [] for imo in tqdm(vessel_imos): vessel_emissions = emissions_api.get_emissions_by_imo(imo, include_eu_emissions=True) vessel_emissions = [emissions.to_dict() for emissions in vessel_emissions] emissions_list.extend(vessel_emissions) emissions_df = pd.json_normalize(emissions_list, sep='') emissions_df.describe()

100%|███████████████████████████████████████████| 27/27 [00:15<00:00,  1.78it/s]

Out[4]:

	IMO	VoyageNumber	VesselTypeID	VesselClassID	Deadweight	Quantity	TransportWorkInMillionTonneMiles	TransportWorkInMillionDwtMiles	EmissionsVoyageCO2InTons	EmissionsVoyageCOInTons	...	EuropeanUnionRegulatedEmissionsPortCallSOxInTons	EuropeanUnionRegulatedEmissionsPortCallPmInTons	EuropeanUnionRegulatedEmissionsStopCO2InTons	EuropeanUnionRegulatedEmissionsStopCOInTons	EuropeanUnionRegulatedEmissionsStopCh4InTons	EuropeanUnionRegulatedEmissionsStopN2OInTons	EuropeanUnionRegulatedEmissionsStopNMVOCInTons	EuropeanUnionRegulatedEmissionsStopNOxInTons	EuropeanUnionRegulatedEmissionsStopSOxInTons	EuropeanUnionRegulatedEmissionsStopPmInTons
count	2.929000e+03	2929.000000	2929.0	2929.000000	2929.000000	2929.000000	2929.000000	2929.000000	2929.000000	2929.000000	...	572.000000	572.000000	626.000000	626.000000	626.000000	626.000000	626.000000	626.000000	626.000000	626.000000
mean	9.562796e+06	126.784227	1.0	87.118129	75800.533971	56108.910891	163.953845	217.178833	1122.464010	0.988094	...	0.102479	0.048364	72.304583	0.062734	0.001359	0.003442	0.069755	1.813836	0.113059	0.036320
std	1.399511e+05	58.187661	0.0	1.004282	28939.074146	23301.761359	236.048850	306.676472	1062.966930	0.935962	...	0.114249	0.030793	202.221402	0.174937	0.003789	0.009510	0.194515	5.099325	0.496942	0.095695
min	9.299111e+06	1.000000	1.0	85.000000	37887.000000	24000.000000	0.118400	0.210267	192.858353	0.169537	...	0.000880	0.006338	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000
25%	9.462354e+06	86.000000	1.0	86.000000	49998.000000	35000.000000	28.452220	40.057118	430.472571	0.378935	...	0.029326	0.014085	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000
50%	9.505819e+06	123.000000	1.0	87.000000	74103.000000	55000.000000	58.804550	82.312392	688.415628	0.606231	...	0.127079	0.061035	15.980352	0.013846	0.000300	0.000782	0.015396	0.422041	0.011419	0.009049
75%	9.688348e+06	163.000000	1.0	88.000000	104649.000000	80000.000000	218.748960	289.765112	1422.189508	1.249819	...	0.127079	0.061035	60.219080	0.052393	0.001135	0.002851	0.058257	1.508939	0.056936	0.029599
max	9.833723e+06	277.000000	1.0	89.000000	149989.000000	148000.000000	3027.411400	3492.910834	10094.160506	8.861095	...	2.443825	0.348220	2584.554489	2.233068	0.048370	0.120924	2.482978	70.337610	9.462714	1.366754

8 rows × 88 columns

We will convert the date columns (StartDate, EndDate) to datetime for easier manipulation. Since we want a yearly analysis, we will assign a "Year" label to each voyage, which we will get from the EndDate.¶

In [5]:

emissions_df['StartDate'] = pd.to_datetime(emissions_df['StartDate'], format='ISO8601')
emissions_df['EndDate'] = pd.to_datetime(emissions_df['EndDate'], format='ISO8601') 
emissions_df['Year'] = emissions_df['EndDate'].dt.year

emissions_df['StartDate'] = pd.to_datetime(emissions_df['StartDate'], format='ISO8601') emissions_df['EndDate'] = pd.to_datetime(emissions_df['EndDate'], format='ISO8601') emissions_df['Year'] = emissions_df['EndDate'].dt.year

To perform our analysis, we group our data per IMO and Year sum the values that we need, i.e. Total CO2 Emissions and EU - regulated CO2 emissions.¶

In [6]:

imos = emissions_df.IMO.unique()

grouped_df = emissions_df.groupby(['IMO', 'Year']).agg(
    VesselName=("VesselName", "first"),
    VesselClass=("VesselClass", "first"),
    Total_CO2_In_Tonnes=("EmissionsVoyageCO2InTons", "sum"), 
    EU_CO2_In_Tonnes=("EuropeanUnionRegulatedEmissionsVoyageCO2InTons", "sum"))#.reset_index()

imos = emissions_df.IMO.unique() grouped_df = emissions_df.groupby(['IMO', 'Year']).agg( VesselName=("VesselName", "first"), VesselClass=("VesselClass", "first"), Total_CO2_In_Tonnes=("EmissionsVoyageCO2InTons", "sum"), EU_CO2_In_Tonnes=("EuropeanUnionRegulatedEmissionsVoyageCO2InTons", "sum"))#.reset_index()

Before we proceed, we will calculate the percentage of EU - regulated emissions per year, as well as the EUAs for the available data.¶

In [7]:

grouped_df['EU_Emissions_Percentage'] = grouped_df['EU_CO2_In_Tonnes']/grouped_df['Total_CO2_In_Tonnes']
grouped_df['EUAs (40%)'] = grouped_df['EU_CO2_In_Tonnes'] * 0.4
grouped_df.head(5)

grouped_df['EU_Emissions_Percentage'] = grouped_df['EU_CO2_In_Tonnes']/grouped_df['Total_CO2_In_Tonnes'] grouped_df['EUAs (40%)'] = grouped_df['EU_CO2_In_Tonnes'] * 0.4 grouped_df.head(5)

Out[7]:

		VesselName	VesselClass	Total_CO2_In_Tonnes	EU_CO2_In_Tonnes	EU_Emissions_Percentage	EUAs (40%)
IMO	Year
9299111	2018	Elka Apollon	Aframax	24487.237390	5356.425477	0.218744	2142.570191
	2019	Elka Apollon	Aframax	12882.690209	3804.605036	0.295327	1521.842015
	2020	Elka Apollon	Aframax	15466.783677	4384.380411	0.283471	1753.752165
	2021	Elka Apollon	Aframax	12755.502182	7164.812591	0.561704	2865.925037
	2022	Elka Apollon	Aframax	22431.696629	2893.248478	0.128980	1157.299391

Finally, we will calculate our estimations using extrapolations from the historical data. We calculate the year to date percentage for the date of the analysis, as well as the the average yearly percentage of EU emissions. Using these two we can estimate the total and EU emissions at the end of the current year, and the number of EUAs needed for the estimated emissions.¶

In [8]:

current_year = 2024

year_to_date_percentage = (dt.datetime.now() - dt.datetime(dt.datetime.now().year,1,1)).days/365

grouped_df['Estimated_end_of_year_CO2'] = np.nan
grouped_df['Estimated_end_of_year_EU_CO2'] = np.nan
grouped_df['Estimated_end_of_year_EUAs'] = 0

for imo in imos:
    average_co2_across_years = grouped_df['Total_CO2_In_Tonnes'][imo].mean()
    average_eu_co2_across_years = grouped_df['EU_CO2_In_Tonnes'][imo].mean()
    percentage_of_avg_eu_co2 = average_eu_co2_across_years/average_co2_across_years
    
    projection_total = grouped_df['Total_CO2_In_Tonnes'][imo][current_year] / year_to_date_percentage
    projection_eu = projection_total * percentage_of_avg_eu_co2
    
    grouped_df.loc[(imo, 2024), 'Estimated_end_of_year_CO2'] = projection_total
    grouped_df.loc[(imo, 2024), 'Estimated_end_of_year_EU_CO2'] = projection_eu
    grouped_df.loc[(imo, 2024), 'Estimated_end_of_year_EUAs'] = round(projection_eu * 0.4)

current_year = 2024 year_to_date_percentage = (dt.datetime.now() - dt.datetime(dt.datetime.now().year,1,1)).days/365 grouped_df['Estimated_end_of_year_CO2'] = np.nan grouped_df['Estimated_end_of_year_EU_CO2'] = np.nan grouped_df['Estimated_end_of_year_EUAs'] = 0 for imo in imos: average_co2_across_years = grouped_df['Total_CO2_In_Tonnes'][imo].mean() average_eu_co2_across_years = grouped_df['EU_CO2_In_Tonnes'][imo].mean() percentage_of_avg_eu_co2 = average_eu_co2_across_years/average_co2_across_years projection_total = grouped_df['Total_CO2_In_Tonnes'][imo][current_year] / year_to_date_percentage projection_eu = projection_total * percentage_of_avg_eu_co2 grouped_df.loc[(imo, 2024), 'Estimated_end_of_year_CO2'] = projection_total grouped_df.loc[(imo, 2024), 'Estimated_end_of_year_EU_CO2'] = projection_eu grouped_df.loc[(imo, 2024), 'Estimated_end_of_year_EUAs'] = round(projection_eu * 0.4)

In [9]:

projected_emissions_for_current_year = grouped_df.loc[grouped_df.index.get_level_values(1)==current_year]
projected_emissions_for_current_year.head(10)

projected_emissions_for_current_year = grouped_df.loc[grouped_df.index.get_level_values(1)==current_year] projected_emissions_for_current_year.head(10)

Out[9]:

		VesselName	VesselClass	Total_CO2_In_Tonnes	EU_CO2_In_Tonnes	EU_Emissions_Percentage	EUAs (40%)	Estimated_end_of_year_CO2	Estimated_end_of_year_EU_CO2	Estimated_end_of_year_EUAs
IMO	Year
9299111	2024	Elka Apollon	Aframax	11856.572240	0.000000	0.000000	0.000000	17105.331492	3432.097933	1373
9314193	2024	Dubai Gold	Panamax	13900.358315	817.786967	0.058832	327.114787	20053.876621	1230.526045	492
9411197	2024	Asahi Princess	Aframax	18612.543976	2740.002717	0.147213	1096.001087	26852.089136	2321.580055	929
9425526	2024	MP MR Tanker 2	MR2	16446.783554	0.000000	0.000000	0.000000	23727.573112	1043.193350	417
9440473	2024	Altesse	Panamax	10742.458029	0.000000	0.000000	0.000000	15498.012571	1324.737525	530
9459096	2024	Nave Orion	MR2	9165.163124	0.000000	0.000000	0.000000	13222.468538	187.828246	75
9461659	2024	Hafnia Pegasus	MR2	14512.583470	0.000000	0.000000	0.000000	20937.126351	1185.199828	474
9462354	2024	Koi	Panamax	9862.682237	0.000000	0.000000	0.000000	14228.770816	0.000000	0
9486922	2024	Alberta	Aframax	11998.141262	2148.311226	0.179054	859.324490	17309.571385	1293.840240	518
9486934	2024	Samos	Aframax	15615.181956	0.000000	0.000000	0.000000	22527.831676	1909.014500	764

To get the total number of EUAs, we just sum the last column.¶

In [10]:

sum(projected_emissions_for_current_year.Estimated_end_of_year_EUAs)

sum(projected_emissions_for_current_year.Estimated_end_of_year_EUAs)

Out[10]:

30229