Cost Models

The cost models used to calculate intervention costs are Excel-based models developed by the RRAP Translation to Deployment team. The models must be requested via QUT (Nick Dendle at QUT).

The compatible versions of the cost models are:

• Coral Aquaculture Deployment : 3.8.0 CA Deployment Model.xlsx

• Coral Aquaculture Production : 3.8.0 CA Production Model.xlsx

Configuration files

Sampling of cost models are dependent on two configuration files. The first is the config.csv file which defines the names, cell positions, and sampling ranges of cost model parameters. A copy of this CSV file is bundled with the package (at cost_eco_model_linker/config.csv).

It is currently not possible to provide a different config.csv file.

An example of this file for the latest compatible cost model versions is below:

Config file for latest model version

cost_type	sheet	factor_names	cell_row	cell_col	range_lower	range_upper	is_cat
production	Dashboard	setupCost	14	12	1000	1000000	FALSE
production	Dashboard	Cost	16	12	1000	1000000	FALSE
production	Dashboard	num_devices	5	5	1000	1000000	TRUE
production	Dashboard	species_no	10	5	19	23	TRUE
production	Conversions	col_spawn_gam_bun	7	6	22500	27500	TRUE
production	Conversions	gam_bun_egg	8	7	10	14	TRUE
production	Conversions	egg_embryo	9	8	0.7	0.9	FALSE
production	Conversions	embryo_freeswim	10	9	0.7	0.9	FALSE
production	Conversions	freeswim_settle	11	10	0.7	0.9	FALSE
production	Conversions	settle_just	12	11	0.7	0.9	FALSE
production	Conversions	just_unit	14	11	6	9	TRUE
production	Conversions	just_mature	14	15	0.7	0.9	FALSE
production	Conversions	1YOEC_yield	19	18	0.7	0.8	FALSE
production	Logistics	optimal_rear_dens	11	5	1	3	TRUE
deployment	Dashboard	setupCost	11	11	1000	1000000	FALSE
deployment	Dashboard	Cost	6	11	1000	1000000	FALSE
deployment	Dashboard	num_devices	5	4	1000	1000000	TRUE
deployment	Dashboard	port	6	4	0	4	TRUE
deployment	Dashboard	DAJ_a_r	9	4	9	10	TRUE
deployment	Dashboard	DAJ_c_s	10	4	1	2	TRUE
deployment	Dashboard	deck_space	13	4	39	40	FALSE
deployment	Lookup Tables	cape_ferg_price	23	7	6000	8000	FALSE
deployment	Lookup Tables	ship_endurance	23	15	13	15	TRUE
deployment	Lookup Tables	distance_from_port	54	8	1	210	FALSE
deployment	Conversions	1YOEC_yield	25	5	0.7	0.8	FALSE
deployment	Logistics	secs_per_dev	84	6	0	1	TRUE
deployment	Logistics	proportion	84	9	0.5	0.55	FALSE
deployment	Logistics	bins_per_tender	83	9	4	6	TRUE
deployment	Logistics	deployment_dur	38	4	25	28	TRUE

At minimum, the file must include the following columns:

• cost_type : the model type the parameter belongs to (currently either production or deployment).

• sheet : the sheet name the parameter occurs on.

• factor_names : a label for the factor.

• cell_row, cell_col: the cell row and column reference for the parameter.

• range_lower, range_upper : the upper and lower bounds for sampling the parameter.

• is_cat : a flag designating whether the parameter is categorical or not.

The config.csv should include info for setupCost and Cost parameters for both the production and deployment models, parameters which are used to extract the setup (CAPEX) and operational (OPEX) cost respectively for a given intervention.

Cost model parameter descriptions

The following table describes the key parameters sampled in the deployment and production cost models. There are many other parameters which could be sampled, but the following were chosen as key sources of uncertainty through consultation with the Translation to Deployment Team and sensitivity analyses.

Parameter name	Description	Expected range	Model
num_devices	The number of devices outplanted in a year. Each device carries 3 baby corals, with a survival rate to 1YO of 1YOEC_yield.	(1000, 5000000)	Deployment and Production
species_no	Number of unique species/regions combinations to be outplanted. Each species and region needs different tanks for production.	(1, 12+)	Production
col_spawn_gam_bun, gam_bun_egg, egg_embryo, embryo_freeswim	Parameters detailing conversions from the spawn to gamete to egg to embryo to freeswimming stage of the coral lifecycle.	See example config file	Production
freeswim_settle, settle_just, just_mature	Parameters detailing conversions from the freeswimming to settlement-ready larvae, to just-settled larva to just-settled unit to mature unit.	See example config file	Production
1YOEC_yield	The number of surviving 1 YO corals per outplanted device.	(0.6, 0.8)	Production and Deployment
optimal_rear_density	The optimal density to rear baby corals at.	(1,3)	Production
port	Index specifying which of port to use, generally overwritten in favour of using distance to port directly.	(0,4)	Deployment
DAJ_a_r	DAJ assembly rate (jig-sec/device).	(9, 10)	Deployment
DAJ_c_s	DAJ count at sea (DAJ/ship)	(1,2)	Deployment
deck_space	Deck space on the ship in m2	Depends on the ship being used	Deployment
cape_ferg_price	Daily rate of the ship being used	Depends on the ship	Deployment
ship_endurance	Number of days the ship can stay out without going back to port	Depends on ship	Deployment
distance_from_port	Distance from port to intervention reef in nautical miles	Depends on intervention reef(s)	Deployment
secs_per_dev	On transect deployment rate of devices	(1,2)	Deployment
proportion	Proportion by which device deployment rate is reduced due to poor visibility	(0.5,0.55), but depends on conditions	Deployment
bins_per_tender	Bins holding devices that can fit in each tender	(4,6) but depends on tender	Deployment
deployment_dur	Days over which the deployment occurs	(25, 28)	Deployment

Sampling the cost models

The Excel-based cost models give CAPEX and OPEX costs of the production and deployment stages of outplanting corals, for a particular input deployment volume, number of species, distance from port to the deployment reef, and other factors, primarily for the ReefMod Engine model. Further details are found in the documentation.

Cost model samples output file

The output files for the cost sampling include 11 cost codes for each intervention year and simulation which are used by CREAM. These are described in the following table:

Cost code	Cost component	Description
1	CAPEX	Sum of setup (capital expenditure) costs for the production and deployment stages.
2	Contingency of CAPEX	% of CAPEX, default is 25%.
3	OPEX	Sum of operational costs for the production and deployment stages.
4	Sustaining capital OPEX	Set to zero for now (assumed to be included in OPEX through the contract).
5	Contingency of OPEX	% of OPEX, default is 25%.
6	Vessel fuel	Only relevant if volunteer vessels are used, set to zero for now.
7	CAPEX - monitoring	Set to zero (assumed no monitoring cost).
8	Contingency CAPEX-monitoring	% of CAPEX monitoring.
9	OPEX - monitoring	Set to zero (assumed no monitoring cost).
10	Sustaining capital OPEX - monitoring	Set to zero (assumed no monitoring cost).
11	Contingency OPEX-monitoring	% of OPEX monitoring

Calculating costs for interventions over multiple years

For outplanting corals over multiple years, the setup costs of production and deployment only scale with the additional corals planned for outplanting compared to the previous year. For example, if 100,000 corals are deployed in the first year, 200,000 in the second year and 200,000 in the third year, the setup costs are $X, $X and $0 respectively. The operational costs are calculated the same way regardless of the year. These year-by-year setup cost differences are not specified in the Excel models, but are handled within the functions cost_calculations.calc_setup_costs() and cost_calculations.calculate_costs().

In particular, for years beyond the first intervention year, the costs are sampled first using the total number of outplants to extract the operational cost. Then the difference in number of corals between the current and previous year is calculated. For differences less than or equal to 0, the setup cost is set to zero. For differences in number of corals greater than zero, the cost is resampled for the difference in number of corals and just the setup cost is updated. This requires double sampling, so interventions which continuously increase the number of outplanted corals over time can take longer to run.

Calculating distance to port for multiple reefs

The Excel-based cost models take as input a single value for distance to port to calculate deployment costs, so it is not specified how to deal with intervening on multiple reefs. For multiple intervention reefs for a single intervention, the total distance to port is calculated as follows:

1. The intervention reefs are clustered into groups of reefs which are a maximum distance
   apart (set by parameter max_dist).

2. The initial distance to port is set as the maximum distance to port for reefs in the
   closest cluster to port.

3. For the remaining clusters, distances are calculated between the furthest reefs from
   port in each cluster.

4. The total distance travelled is then estimated as the initial distance to the closest
   cluster’s furthest reef plus the remaining distances between the furthest reefs from
   port in each of the remaining clusters, travelling in order from closest cluster to
   port to furthest cluster from port.