Case Study. Architectural Flexibility in Commercial Communication Satellite Fleets. Reference

Transcription

1 Case Study Architectural Flexibility in Commercial Communication Satellite Fleets Reference Value-at-Risk Analysis for Real Options in Complex Engineered Systems, Hassan, R., de Neufville, R., de Weck, O., Hastings, D. and McKinnon, D., IEEE Conference on Large-Scale Infrastructures, Hawaii, 005

2 Fleet Architecting Framework Objective: to integrate spacecraft engineering design with economic analysis for the purpose of maximizing the financial value of a fleet to the operator under uncertain, evolving market conditions. Economic Domain Feedback for payload-spacecraft feasibility Demand Distributions (dynamic) Payload(s) Requirement Specification Technical Domain ROV Revenue Cost E(NPV ) S/C Sizing & Reliability Models Optimization Algorithm (GA) 3 Dynamic Demand Uncertainty i Hypothetical case investigating demand forecasts in areas similar in size to Continental Europe and South America i Coverage Area I Demand for a Ku-Band service Coverage Area: 40 deg Expected initial bandwidth capacity required : 00 W TWTAs Demand decreases over the satellite lifetime (demographics, competition, etc..) i Coverage Area II Demand for the same Ku-Band service is estimated to start in 5 years Coverage Area: 5 deg Expected initial bandwidth capacity required : 5 00 W TWTAs Demand increases over the satellite lifetime (market matures) 4

3 3 Dynamic Demand Uncertainty i Hypothetical case investigating demand forecasts in areas similar in size to Continental Europe and South America i Coverage Area I Demand for a Ku-Band service Coverage Area: 40 deg Expected initial bandwidth capacity required : 00 W TWTAs Demand decreases over the satellite lifetime (demographics, competition, etc..) i Coverage Area II Demand for the same Ku-Band service is estimated to start in 5 years Coverage Area: 5 deg Expected initial bandwidth capacity required : 5 00 W TWTAs Demand increases over the satellite lifetime (market matures) 5 Dynamic Demand Uncertainty CA-I Forecast Transponder Demand CA-II Forecast Transponder Demand Probability Probability Number of Transponders 0 Number of Transponders Y:Y Y6:Y Y:Y Y6:Y Y:Y Y6:Y Y:Y Y6:Y Forecast transponder demand distributions for a fixed satellite service in two hypothetical markets over 0 years 6 3

4 4 Economic Models i Discount Rate 0% i Revenue Transponder Price $,50,000 Annual Price Decrease 4% i Cost Satellite $,000,000 x No. of Transponders + Bus Cost (see table) Launch (see table) Insurance 0% of (Satellite Cost + Launch Cost) Ops + Marketing + Admin $ 5,000,000 / year for a large satellite $ 3,500,000 / year for a small satellite $ 500,000 / year for co-located satellites Payload Power (kw) Bus Cost 4 $ 35,000,000 7 $ 50,000, $ 70,000,000 Satellite Mass (kg) Launch Cost,000 3,500 $ 90,000,000 3,500 5,000 $ 0,000,000 5,000 < $ 0,000,000 7 Architectural Configuration of Fleet Solutions i Technical flexibility levers: Number of spacecraft Payload size onboard each spacecraft Payload flexibility switch Timing or relative sequencing of deployment Architectural Parameters Rigid Fleet Fleet I Fleet II Fleet III Number of spacecraft Active transponders S/C I in CA-I: S/C II in CA-II: S/C I: S/C II: S/C I: S/C II: Payload flexibility none yes both yes both yes Deployment stage S/C I in CA-I: stage I S/C II in CA-II: stage II stage I S/C I: stage I S/C II: stage I S/C I: stage I S/C II: stage II 8 4

5 5 Parameters of Optimal Spacecraft Designs Rigid Fleet Fleet I Fleet II Fleet III Optimized Design Parameter S/C I in CA-I S/C II in CA-II S/C S/C I S/C II S/C I S/C II Total Launch Mass (kg) 4,54 4,888 5,96,75,75,75,75 Payload Active HPAs* Available HPAs HPA efficiency (%) Payload Power (W) 6,9 6,9 7, Bus Solar Array Area (m ) Battery Mass (kg) Radiator Area (m ) Real Options Valuation Results Probability Rigid Fleet - CDF Rigid Fleet - Mean Flex Fleet I - CDF Flex Fleet I - Mean Flex Fleet II - CDF Flex Fleet II - Mean Flex Fleet III - CDF Flex Fleet III - Mean Net Present Value ($ million) Value-at-Risk-and Gain (VARG) Distributions for Fleet Architectures 0 5

6 6 Economic Values of Fleet Architectures Architectural Value Parameter ($ million) Rigid Fleet Fleet I Fleet II Fleet III E(NPV) Std(NPV) Flexibility Value Fixed cost, pay year Fixed cost, pay year PV(fixed cost) at year Maximum possible gain Maximum possible loss Design Philosophy Take-aways i Ultimate Objective of Design is to provide value. Technical excellence is a means to this end, not end in itself. i Fixed design cannot adjust to evolving situations, thus may perform badly in unfavorable cases (see value-at-risk-and-gain curves). i designs can be adjusted to conditions that develop, avoiding downside results and exploiting upside opportunities i At the least, we can reduce the range of the possible economic outcomes, and thus reduce the necessary risk premium (via CAPM). i In general we can also increase Expected Value of System, sometimes dramatically. 6