Select input variables for sensitivity analysis

[bwbush]

For which of the constants listed below do we want to study sensitivities? We have sufficient computational resources to study a hundred or so of the ~250 variables. Please check the box next to the variable to mark it for inclusion.

Constants:

Developer Decisions

• [x] advertising budget
• [x] advertising start time
• [x] base external investor ask rate
• [x] business as usual runway
  • Captures how conservative the developers are
• [x] commercial plant capacity
• [x] commercial plant capacity input
• [x] commercial plant startup period
• [x] demo capacity
  • I think we should pick scales for the pilot and demo plants as fractions of the planned commercial plant, and calculate accordingly - so we only do sensitivity on one capacity variable but three variables response. Capacity doesn't influence start-up, failures or construction time so all it would do is change the capital and operating expenses for piloting and demoing.
• [x] demoing acceptable rate
• [x] demoing failure default recovery time
• [x] demoing failure distribution max
• [x] demoing management overhead
  • This one and piloting management overhead capture the importance of conserving funds to run a plant as opposed to funding management, and also can have negative impacts on external investor asks.
• [x] management response time
• [x] max management runway response
• [x] minimum runway
  • Also captures how conservative the developers are, along with business as usual runway
• [x] pilot and demo response time
• [x] pilot capacity
  • See commend on demo capacity
• [x] piloting acceptable rate
• [x] piloting failure default recovery time
• [x] piloting failure distribution max
• [x] piloting management overhead
• [x] regulatory process starting point
• [x] startup demoing period
  • The four startup [scale] period/rate parameters will have some interesting interactions but all influence how long it takes to start drawing down the Continuous [scale] Hours stocks. Maybe we should fix either periods or rates and then vary the others.
• [x] startup demoing rate
• [x] startup piloting period
• [x] startup piloting rate
• [x] target demo hours
• [x] target pilot hours

External and Internal Investor Decisions

• [x] aversion to NPV deviation
• [x] expected continuity of government policy
• [x] investor history
• [x] investor optimism
• [x] number of missed stagegates allowed
• [x] payback period multiplier
• [x] required internal return
• [x] required return multiplier
• [x] stagegate length
• [x] strategic value to external investors

Government Decisions

• [x] government capital cost share
• [x] government operating cost share
• [x] government operating grant period
• [x] government production incentive
• [x] government research cost share

Market Characteristics

• [x] bioproduct offtake agreement
• [x] duration of intellectual property
  • This one and effectiveness of intellectual property influence the NPV so should probably be included
• [x] effectiveness of intellectual property
• [x] elasticity of demand
• [x] elasticity of supply
• [x] expected green premium
• [x] initial market size
• [x] long term market size
• [x] market growth rate

Bioproduct Characteristics

• [x] commercial capital cost input
• [x] commercial fixed operating cost input
• [x] commercial process yield input
• [x] commercial variable operating cost input
• [x] bioproduct long term price
• [x] bioproduct performance advantage
• [x] bioproduct price fluctuations
• [x] bioproduct price reversion time
• [x] custom feedstock long term price change
• [x] custom feedstock maximum fluctuation magnitude
• [x] custom feedstock periodic fluctuation magnitude
• [x] custom feedstock reversion time
• [x] custom feedstock starting price
• [x] feedstock approval maximum cost
• [x] feedstock approval maximum time
• [x] feedstock approval required switch
• [x] feedstock selector array[custom]
• [x] incumbent long term price trend
• [x] incumbent market share target economic
• [x] incumbent maximum fluctuation
• [x] incumbent price response magnitude
• [x] incumbent reversion time
• [x] incumbent starting price
• [x] pathway approval maximum cost
• [x] pathway approval maximum time
• [x] pathway approval required switch
• [x] product approval maximum cost
• [x] product approval maximum time
• [x] product approval required switch

Misc / Discuss before Categorizing

• [x] random stream

• [x] researching impact on demoing

  • The effect of this and researching impact on piloting could be difficult to interpret and I'm not sure how controllable they would be - maybe include?

• [x] researching impact on piloting

• [x] retrofit delay

Excluded variables

• abandonment rate Difficult to control, but could maybe construct some scenarios along with advertising budget
• advertising effectiveness It would likely be more interesting to vary the advertising amount, since that's what developers have direct control over
• biofuel feedstock[corn]
• biofuel feedstock[corn stover]
• biofuel feedstock[soybean]
• biofuel feedstock[commodity sugar]
• biofuel feedstock[generic perennial]
• biofuel feedstock[generic fossil]
• biofuel feedstock[custom]
• biofuel technology unit operations[onsite lignocellulosic biomass processing]
• biofuel technology unit operations[onsite starch biomass processing]
• biofuel technology unit operations[lignocellulosic pretreatment and conditioning]
• biofuel technology unit operations[starch biomass fermentation]
• biofuel technology unit operations[sugar fermentation]
• biofuel technology unit operations[fixed bed reactor]
• biofuel technology unit operations[enzymatic hydrolysis]
• biofuel technology unit operations[flash tank]
• biofuel technology unit operations[solids liquids separation]
• biofuel technology unit operations[distillation]
• biofuel technology unit operations[hydrolysate clarification]
• biofuel technology unit operations[agitated pressurized reactor]
• biofuel technology unit operations[biological conversion]
• biofuel technology unit operations[crystallizer]
• biofuel technology unit operations[catalytic conversion]
• biofuel technology unit operations[centrifuge]
• biofuel technology unit operations[extraction purification and fuel finishing]
• biofuel technology unit operations[extraction purification and product finishing]
• bioproduct technology unit operations[onsite lignocellulosic biomass processing]
• bioproduct technology unit operations[onsite starch biomass processing]
• bioproduct technology unit operations[lignocellulosic pretreatment and conditioning]
• bioproduct technology unit operations[starch biomass fermentation]
• bioproduct technology unit operations[sugar fermentation]
• bioproduct technology unit operations[fixed bed reactor]
• bioproduct technology unit operations[enzymatic hydrolysis]
• bioproduct technology unit operations[flash tank]
• bioproduct technology unit operations[solids liquids separation]
• bioproduct technology unit operations[distillation]
• bioproduct technology unit operations[hydrolysate clarification]
• bioproduct technology unit operations[agitated pressurized reactor]
• bioproduct technology unit operations[biological conversion]
• bioproduct technology unit operations[crystallizer]
• bioproduct technology unit operations[catalytic conversion]
• bioproduct technology unit operations[centrifuge]
• bioproduct technology unit operations[extraction purification and fuel finishing]
• bioproduct technology unit operations[extraction purification and product finishing]
• capital cost maturity factor
• commercial construction time - This time is fixed and there's no logic for unanticipated delays or construction failures
• commodity sugar long term price change
• commodity sugar maximum fluctuation magnitude
• commodity sugar reversion time
• commodity sugar starting price
• corn long term price change
• corn maximum fluctuation magnitude
• corn periodic fluctuation magnitude
• corn reversion time
• corn starting price
• corn stover long term price change
• corn stover maximum fluctuation magnitude
• corn stover periodic fluctuation magnitude
• corn stover reversion time
• corn stover starting price
• demo capital cost scaling factor
• demo construction time - See commercial construction time comment
• demo fixed operating cost scaling factor
• demoing delta TRL
• feedstock approval minimum cost The minimum regulatory costs are already set very low and I don't think varying them will be interesting
• feedstock approval minimum time - See comment on feedstock approval minimum cost
• feedstock selector array[corn]
• feedstock selector array[corn stover]
• feedstock selector array[soybean]
• feedstock selector array[commodity sugar]
• feedstock selector array[generic perennial]
• feedstock selector array[generic fossil]
• fixed operating cost maturity factor
• fossil feedstock long term price change
• fossil feedstock maximum fluctuation magnitude
• fossil feedstock reversion time
• fossil feedstock starting price
• include demo in cumulative production I don't think changing this variable will affect any of the model results (?)
• initial TRL
• IS carry forward period
• IS equipment useful lifetime
• IS tax rate
• pathway approval minimum cost See comment on feedstock approval minimum cost
• pathway approval minimum time See comment on feedstock approval minimum cost
• perennial long term price change
• perennial maximum fluctuation magnitude
• perennial periodic fluctuation magnitude
• perennial reversion time
• perennial starting price
• pilot and demo runway response
• pilot capital cost scaling factor
• pilot construction time - See commercial construction time comment
• pilot fixed operating cost scaling factor
• piloting delta TRL
• process yield maturity factor
• product approval minimum cost See comment on feedstock approval minimum cost
• product approval minimum time See comment on feedstock approval minimum cost
• research response time Unlikely to have a big impact
• researcher average salary
• researching management overhead Researching spending doesn't have a huge impact on overall techno-economics
• seed investment Provided this value is not zero, changing it probably wouldn't have an impact on results - unfavorable economics won't be significantly improved using just the seed (need more funds for piloting)
• shared learning rates for feedstock logistics[corn]
• shared learning rates for feedstock logistics[corn stover]
• shared learning rates for feedstock logistics[soybean]
• shared learning rates for feedstock logistics[commodity sugar]
• shared learning rates for feedstock logistics[generic perennial]
• shared learning rates for feedstock logistics[generic fossil]
• shared learning rates for feedstock logistics[custom]
• shared learning rates for unit operations[onsite lignocellulosic biomass processing]
• shared learning rates for unit operations[onsite starch biomass processing]
• shared learning rates for unit operations[lignocellulosic pretreatment and conditioning]
• shared learning rates for unit operations[starch biomass fermentation]
• shared learning rates for unit operations[sugar fermentation]
• shared learning rates for unit operations[fixed bed reactor]
• shared learning rates for unit operations[enzymatic hydrolysis]
• shared learning rates for unit operations[flash tank]
• shared learning rates for unit operations[solids liquids separation]
• shared learning rates for unit operations[distillation]
• shared learning rates for unit operations[hydrolysate clarification]
• shared learning rates for unit operations[agitated pressurized reactor]
• shared learning rates for unit operations[biological conversion]
• shared learning rates for unit operations[crystallizer]
• shared learning rates for unit operations[catalytic conversion]
• shared learning rates for unit operations[centrifuge]
• shared learning rates for unit operations[extraction purification and fuel finishing]
• shared learning rates for unit operations[extraction purification and product finishing]
• shared learning threshold for bioproduct production
• shared learning threshold for feedstock logistics
• soybean long term price change
• soybean maximum fluctuation magnitude
• soybean periodic fluctuation magnitude
• soybean reversion time
• soybean starting price
• staff sent to demo
• staff sent to pilot
• total staff
• variable operating cost maturity factor
• word of mouth contact rate See comment below on word of mouth effectiveness
• word of mouth effectiveness This parameter can't really be controlled but could maybe be used to represent how enthusiastic the market is about bioproducts?

Graphical functions:

• required payback period trend
• required return trend
• supply to demand ratio price impact

[rjhanes]

Questions & comments for discussion:

• Does the feedstock selector array count as multiple variables or one variable vector?
• I'm inclined for this analysis to exclusively use the custom feedstock price model instead of switching between the pre-built ones - I think this will mean fewer variables to vary overall and might make results analysis easier?

I checked variables that I think should definitely be included.

I've crossed out variables that I think should be excluded:

• all shared learning-related variables. It looks like there's a lot of these but that's due to arrayed variables. I think this is no longer a topic of interest to the client and if it is, we can easily post-process the model results we're interested in (success/failure, production, etc) to get information on shared learning.
• the maturity factor variables. These are artifacts from an older version of the model, aren't currently used, and weren't visible on the Vensim view.
• the three graphical functions. IIRC there's no way to specify these using .cin files, so we'd have to vary them manually.
• total staff, staff sent to demo, staff sent to pilot, researcher average salary. Unless we're looking at an extremely small scale process, staff salaries comprise a very small portion of overall expenses.
• piloting delta TRL and demo delta TRL - these should probably remain fixed and won't affect results
• initial TRL - same reasoning
• all of the "scaling fator" parameters. These are modeling/calculation assumptions; varying them doesn't represent different scenarios.
• all of the pre-built price model variables - see second comment above. I'm assuming we use the custom feedstock price model throughout our analysis, but this can change easily.
• IS carry forward period, IS tax rate, IS equipment useful lifetime - similarly to the scaling factors, these are more modeler assumptions than scenario parameters

Any variables not checked or crossed out I'm either ambivalent about or don't think they'd add a lot of value to our analysis.

[rjhanes]

I reduced the list down to 85 variables for inclusion and included justifications for most of the additional exclusions.

[Lauren-Sittler]

I have looked over the list and Rebecca's comments. Almost every variable I would include from the views I documented is already included. I only have questions regarding:

-probability of successful investor ask -probability of internal project success -expected technology readiness level progress

None of these are input variables; we have assumed them to be functions of other variables. However, the functional forms we used are very generic. They were derived based on data from other industries. I'm wondering if it would make sense to add multipliers that we could then include in the sensitivity analysis.

[bwbush]

Re feedstock selector array, I think we should keep with the custom feedstock for now.

I've incorporated these variables into the elementary-effects screening. I suggest we close this issue and create new issues if we decide to add or remove input variables for the sensitivity analysis.

[rjhanes]

I've checked all the parameter ranges and adjusted where necessary to prevent errors at extreme values and created a template input file in a6389f7