Business Case Analysis of the Towed Glider Air Launched System (TGALS)

Darryl. W. Webb1, Mclinton B. Nguyen,2 Robert W. Seibold,3 and Frank C. Wong4 The Aerospace Corporation, El Segundo, CA 90245-4609

and

Gerald D. (Jerry) Budd5 National Aeronautics and Space Administration, Armstrong Flight Research Center, Edwards, CA 93523-0273

The Aerospace Corporation developed an integrated Business Case Analysis (BCA) model on behalf of the NASA Armstrong Flight Research Center (AFRC). This model evaluated the potential profitability of the Towed Glider Air Launched System (TGALS) concept, under development at AFRC, identifying potential technical, programmatic, and business decisions that could improve its business viability. The model addressed system performance metrics; development, production and operation cost estimates; market size and product/service positioning; pricing alternatives; and market share. Projected annual expenses were subtracted from projected annual revenues to calculate cash flow, return on investment (ROI), and net present value (NPV). This comprehensive model included parametric cost predictions for this system’s development, production and operation, as well as adjustments for organizational complexity associated with integrating TGALS flight and ground components.

1 Senior Project Leader, Economic Market and Analysis Center, 2310 E. El Segundo Blvd., MS M4/034. 2 Engineering Specialist, Economic Market and Analysis Center, 2310 E. El Segundo Blvd., MS M4/034. 3 Senior Project Leader, Civil & Commercial Launch Projects, 2310 E. El Segundo Blvd, MS M1/132, AIAA Associate Fellow. 4 Director, Economic Market and Analysis Center, 2310 E. El Segundo Blvd., MS M4/034. 5 Air Launch Development Project Manager, Advanced Planning and Partnerships Office, P.O. Box 273, MS 2701.

1 American Institute of Aeronautics and Astronautics Nomenclature

A-SOA = Advanced State-of-the-Art AFRC = Armstrong Flight Research Center BCA = Business Case Analysis CAD/CAM/CAE = Computer Aided Design, Manufacturing, and Engineering ICAM = Industrial Capability Assessment Model IM&P = Improvements in Methods and Processes IRR = Internal Rate of Return LEO = Low Earth Orbit LV = Launch Vehicle NASA = National Aeronautics and Space Administration NPV = Net Present Value PBP = Payback Period PIR = Performance Improvement Rate ROI = Return on Investment SME = Subject Matter Expert SOA = State-of-the-Art TGALS = Towed Glider Air Launched System TRL = Technology Readiness Level WBS = Work Breakdown Structure

I. Introduction T he integrated TGALS Business Case Analysis (BCA) model identifies technical, programmatic and business factors that influence the commercial feasibility of the TGALS concept. The model includes parametric cost estimates for developing, integrating, testing, and producing key flight system and ground system components and associated equipment, as well as operational and maintenance costs of these components over the system’s life. Cost projections were developed for different launch rates and maximum payload sizes (up to 600 kg). These estimates include both demonstration and operational vehicles. Business feasibility is assessed in terms of cash flows, Return on Investment (ROI), payback period, and Net Present Value (NPV), using both a traditional acquisition scheme as well as a funding mechanism proposed by NASA-AFRC. To improve objectivity, program independent historical data and analyses were used whenever possible to benchmark calculated financial metrics and program estimated values and to provide sensitivity analysis for the differences identified.

II. Analytic Framework

A. TGALS System Components The TGALS system includes a modified tow aircraft, glider, glider rocket motor, launch vehicle (LV), and ground infrastructure. The glider with under-carried LV is shown in Figure 1. This concept has many advantages:

 High Performance . Lifting of significantly larger geometric payloads to altitude vs. modifying a same-size business jet or commercial transport aircraft to carry the launch vehicle externally . Ability to release the LV at an elevated flight path angle and high altitude coincident with the optimal trajectory for the LV . Up to 70% increase in payload weight to orbit (vs. ground launch), 30% increase due to releasing the LV at an elevated flight path angle by the glider performing a “pull-up” maneuver prior to LV release (Figure 2).  Agility . Low overhead flight operations allows for rapid turn-around to launch satellites quickly . No dependence on critical ground-based launch facilities/assets  Safety & Mission Assurance . Unmanned glider eliminates human safety concerns for carrying LV . Restartable sustainer motor extends glide home distance following launch or abort . Glider capable of landing with LV attached in event of mission abort

2 American Institute of Aeronautics and Astronautics  Flexibility . Glider plug-and-play center wing allows multiple simultaneous LV build-ups . Inexpensive gliders can be staged at nearly any airfield, increasing launch opportunities . Tow plane can be a slightly modified existing aircraft that simply adds towed launch to duties . Glider concept is scalable from very small to larger LVs

Figure 1. TGALS Glider and Launch Vehicle

Figure 2. Pull-up Maneuver

B. BCA Architecture An overview of the Aerospace Corporation developed solution for the TGALS BCA is shown in Figure 3. The solution consists of linking several models into a decision framework. Subject Matter Expert (SME) inputs were used

3 American Institute of Aeronautics and Astronautics to customize Aerospace’s Industrial Capability Assessment Model (ICAM), which is a technology-based cost model (upper left of Figure 3).

Tourist Market Capture Rate 100

Tow Aircraft Glider Liquid Rocket Launch Vehicle Aircraft Avionics & Launches of 1-50 kg Satellite 90 Forecast Tow Cable Tow Retract Airframe Avionics SW Motor Fuel Tanks Fuel Sys 1st Stg 2nd Stg 3rd Stg 4th Stg Fairing Stg Str Stage Cabling Avionics Design Cycle Experience (5) Mods Power High = 80% M Mode (0=str,1=el) 0.000 0.000 0.000 0.000 1.000 2.000 0.000 0.000 0.000 0.500 0.000 0.000 0.000 0.000 0.000 0.000 0.000 1.000 10000 Reliability & Safety (4) MS Market Segment 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 by 2024 100 y = 7E-233e0.2676x 80 Low = 80% IM&P - General (1) Performance (7) AL Assembly level 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 PPD Prime profit Dev 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% R² = 0.8498 11-50 kg typically by 2032 Cost of Technology (8) PPP Prime Profit Prod 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 90 CAD,CAM,CAE,IDS (2) Buy Buy % 100% 100% 100% 80% 80% 35% 100% 100% 100% 80% 100% 100% 100% 100% 100% 100% 100% 80% manifested 2 per launch YrDH Econ Yr $ 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 y = 36440ln(x) - 277055 70 Physical Characteristics (3) Criticality of Failure (6) R² = 0.7651 80 IMP IM&P Yr 1997 2005 2000 1990 2005 2016 2000 2016 2016 2010 2016 2016 2016 2016 2016 2016 2010 2010 1000 Pyr Perf Tech Yr 1997 2005 2000 1990 2005 2016 2000 2016 2016 2010 2016 2016 2016 2016 2016 2016 2010 2010 CAD CAD CAM Yr 1997 2005 2000 1990 2005 2016 2000 2016 2016 2010 2016 2016 2016 2016 2016 2016 2010 2010 60 70 EPCE Eng,PM Cycle Exp yr 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 Program QNA Qty NHA 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 DP Protos 1.00 3.00 3.00 2.00 2.00 1.00 1.00 2.00 2.00 2.00 8.00 8.00 8.00 8.00 8.00 4.00 4.00 4.00 Market Start 60 PBQ Planned Build Qty 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Capture PQ Prior Qty 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 50 100 BQ Prod Build Qty 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Rate 50 AR Annual Rate 1 1 1 1 1 1 1 1 1 1 8 8 8 8 8 8 8 8 HR HR HR NHR NHR NHR NHR NHR NHR NHR NHR NHR NHR NHR NHR NHR NHR NHR % ECF ECF Air - Air - Supersonic - Supersonic - Supersonic - Supersonic - Design Air - Commercial Air - Commercial Air - Commercial Air - Commercial Non-HR, SL Non-HR, SL Non-HR, SL Non-HR, SL Non-HR, SL Non-HR, SL Non-HR, SL Non-HR, SL 40 Commercial Commercial Commercial Commercial Commercial Commercial 1-10 kg typically High = 30% 40 IOC Produceability (9) Satellites Launched S Size (kg) or SLOC 400 600 300 3,000 200 0 9 56 26 17 6,229 2,008 689 146 1,850 925 8 50 manifested 5 per launch

Des Design Complexity 0.59 0.12 0.55 1.29 0.82 0.55 0.20 0.50 1.00 1.00 0.41 0.41 0.41 0.41 0.35 0.35 0.35 0.35 by 2020 Expenditures $ . $ Expenditures 30 ND New Design 20% 31% 31% 100% 0% 100% 6% 100% 100% 75% 31% 31% 31% 31% 31% 31% 31% 8% 30 R Redund % 0% 17% 17% 0% 0% 0% 0% 25% 0% 0% 0% 0% 0% 0% 0% 0% 17% 17% 10 Acceleration Costs 0.1464x INT Integration (internal) 1.0 3.0 3.0 4.0 5.0 5.0 4.0 4.0 5.0 5.0 6.0 6.0 6.0 6.0 6.0 6.0 11.0 11.0 y = 1E-127e Produceability Epsilon-1 (cplx) 0.047 0.001 0.002 0.008 0.019 0.273 0.089 0.003 0.003 0.009 0.006 0.006 0.009 0.019 0.001 0.001 0.004 0.014 R² = 0.5864 20 Produceability End PIR % = 0.5% 0.5% 0.5% 0.5% 5.3% 10.1% 0.5% 0.5% 0.5% 2.9% 0.5% 0.5% 0.5% 0.5% 0.5% 0.5% 0.5% 5.3% 20 Normal Funding CIC % 87% 87% 87% 87% 87% 87% 93% 93% 93% 93% 92% 92% 92% 92% 92% 92% 92% 92% y = 4280.6ln(x) - 32542 10 Cum mults 1.365 1.445 1.395 1.210 2.145 1.106 1.395 1.562 1.562 1.568 1.562 1.562 1.562 1.562 1.562 1.562 1.497 1.760 R² = 0.667 T-1 4.801 0.130 0.134 3.232 1.826 0.000 0.609 0.089 0.048 0.110 4.891 2.196 1.569 1.072 0.312 0.133 0.028 0.405 Low = 6% NRCC NonRec Cost Calib 0.50 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 3.00 3.00 3.00 3.00 1.00 1.00 1.00 1.00 1 10 by 2020 0 RCC Prod Cost Calib 0.25 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 NRC NonRec Cost $m 0.627 0.007 0.037 11.331 - 0.001 0.014 0.040 0.057 0.238 2.220 0.997 0.712 0.487 0.040 0.017 0.004 0.039 2000 2005 2010 2015 2020 2025 2030 2035 PRC Prototype Cost $m 1.200 0.313 0.322 5.623 3.176 0.000 0.609 0.165 0.089 0.204 30.467 13.681 9.773 6.680 1.942 0.449 0.096 1.371 Year DC Total Dev Cost $m 1.827 0.320 0.360 16.954 3.176 0.002 0.623 0.205 0.146 0.442 32.687 14.678 10.485 7.166 1.983 0.466 0.100 1.409 0 RC Production Cost $m 1.200 0.130 0.134 3.232 1.826 0.000 0.609 0.089 0.048 0.110 4.891 2.196 1.569 1.072 0.312 0.133 0.028 0.405 TC Total Predicted $m 3.027 0.450 0.493 20.185 5.002 0.002 1.231 0.293 0.194 0.552 37.578 16.874 12.054 8.239 2.295 0.598 0.128 1.814 2010 2015 2020 2025 2030 2035 2040 Program Months PAUC Prod auc $m 1.200 0.130 0.134 3.232 1.826 0.000 0.609 0.089 0.048 0.110 4.891 2.196 1.569 1.072 0.312 0.133 0.028 0.405 Year From Design Initiation Alternative payload configurations Market forecasting Competitive assessment Technology forecasting Price demand elasticity Industry capability assessment Specific system components Profit application feasibility Design difficulty and inherited development Competition effects Capture rate prediction Profit mark-up predictions

Management Decision Framework Value Utilities Uncertainty, Sensitivity, Risk

1 5.00 Excellent Cumulative Financials 4.00 Good Cost Value Function 3.00 Marginal 2,500 2.00 Poor PreTax Profit 1.00 Inadequate 1.00

Weighted Scores Utility Scoring 0.90 2,000 Cum Revenue Case 1 2 3 4 5 6 0.80 Candidate 1 Candidate 2 Candidate 3 Candidate 4 Candidate 5 Candidate 6 Cum Expenses Wtg High 0.70 1,500 100% Composite 3.37 2.50 3.68 3.19 3.97 2.04 0.60 Low

0.50 Value 20% ROI 5.00 3.00 3.50 5.00 4.50 2.00 0.40 1,000

5% NPV 1.00 4.00 4.50 1.00 4.50 1.00 0.30 20% Breakeven 4.00 2.00 2.50 4.00 5.00 4.00 0.20 in $ millions 2016 500 5% Capture Rate 0.50 4.00 4.50 0.50 3.00 0.50 0.10 0% Competitive Posture

0% Market Health - - 10,000 20,000 30,000 40,000 0 10% Safety 4.00 3.50 4.00 2.00 4.00 3.60 2015 2020 2025 2030 2035 2040 2045 2050 5% Organization 2.00 2.00 2.50 2.00 3.00 2.00

10% Performance -1 5.00 2.50 3.00 5.00 5.00 2.30 Cost -500 10% Performance -2 3.00 2.50 4.00 3.00 3.50 3.00 Year

5% Business Risk 4.20 1.00 4.00 4.20 3.50 1.50

5% Technical Risk 4.20 1.00 4.00 4.20 3.50 1.50 5% Risk Mitigation 4.20 2.00 4.00 4.20 4.20 1.00

Figure 3. BCA Modeling Architecture

For the model, the TGALS performance capability is assessed against customer decision factors and LV market forecasts. Market capture rate forecasts are based upon market demand, price, competition, and price elasticity (upper right of Figure 3). Time phased expense and revenue streams are used to calculate financial metrics such as ROI and NPV. Monte Carlo simulations are used to quantify risk (effects of input uncertainty on the modeling results). Optional value functions are used to adjust weighting factors for TGALS system characteristics (i.e., resilience, price, launch convenience). In the lower left of Figure 3, several candidate configurations are scored for desirability. A rollup of metrics gives a composite score for each system candidate in the management decision framework. Business case desirability is improved through optimizing the model’s output through consecutive iterations.

C. Capturing Advantages of the TGALS Concept The unique and technologically diverse characteristics of TGALS cover a wide range of legacy systems that take advantage of the reusability and efficient flight physics of gliders (Figure 4). An existing aircraft is modified for towing, becoming a human-rated reusable first stage. The glider and the sustainer rocket motor is a new, non-human- rated and reusable second stage. The launch vehicle is ignited in a steep climb above 40,000 feet altitude in lieu of a ground launch. Because the takeoff is from an airport without a launch pad, the ground infrastructure required is minimal. The tow aircraft is a modified business jet or commercial transport for larger payloads, either one with minimal modifications using mature technologies. Components of the modeling system chosen by Aerospace are designed to capture the advantages of TGALS concepts, viz., (1) reusability, (2) technological maturity, (3) existing designs with minimal modifications, and (4) human and non-human rating.

4 American Institute of Aeronautics and Astronautics Weight distributed across 2 airframes

Lift Lift Pull Pull Thrust Drag Drag Weight Weight Launch Vehicle Engines Fuel Avionics Hydraulics Crew Simplified force vectors Sustainer Motor Life Support Avionics

Figure 4. TGALS Force Vectors

III. Major Modeling Components

A. Industrial Capability Assessment Model (ICAM) The ICAM1 hardware cost prediction model is an Aerospace-built parametric system based upon the movement of technology over time correlated with concomitant costs. The primary drivers are described below: 1. CAD/CAM/CAE – Based on the year of the version of Computer Aided Design (CAD). Design costs are reduced as later versions of CAD are utilized. Older legacy systems might initially result in reduced design costs. However, these systems are more expensive to modify and maintain (Figure 5). 2. IM&P – Improvements in Methods and Processes – Based on the year the equipment is developed. It represents the general processes and methods of inventory control, machining, material process sophistication and assembly, and test procedures (Figure 5). 3. Producibility – Describes the amount of time dedicated to mature the manufacturability of a product, reducing complexity, mass, and the number of parts and therefore cost (Figure 5). 4. Design Cycle Experience – The year indicating the experience level of program management and systems engineers. This is the number of programs they have experienced from beginning to end (i.e., program cycle experience). Present-day engineers tend to have experienced fewer programs than engineers working in the 1960’s and 1970’s. Having fewer program experiences results in needing more hours to finish tasks, less insight for decision making, and potentially increased costs (Figure 5). 5. Performance Improvement Rate (PIR) – The rate of technology increase of the equipment on an annual compounding basis. Chosen from a library of hundreds of ‘Moore’s Laws (Figure 5). 6. Cost of Technology – The cost of increasing technology over time expressed as an annual compounding percentage. All things equal, the cost of technology increases but at a slower pace than the PIR thus yielding greater capability per dollar than older technologies (Figure 5). 7. Design Complexity – A table value that describes the difficulty of the engineering task ranging from simple engineering support to difficult advancement of the state-of-the-art (SOA) above the normal technology path. 8. New Design – The amount of a normally complete engineering task that needs to be accomplished expressed as a percentage. 9. Internal and External Integration – These values are used to reflect the difficulty of integrating elements of the equipment. 10. Prototypes – The amount of development and test hardware including simulators, engineering developmental models, prototypes, and qualification units. 11. Planned Build Quantity – The number of production units planned to be built. As planned build increases so does development cost due to producibility and quality concerns. Planned build affects the production T-1

5 American Institute of Aeronautics and Astronautics value and the cost quantity improvement curve. For example, very high planned build rates simplify the design, set the tooling rate, and increase automation, therefore flattening the production rate curve. 12. Production Quantity – The current build quantity needed.

Figure 5. Multidimensional Technology Applied to Cost

The above variables are considered together and are highly interdependent. For example, a low new-design value for an older legacy system is not consistent with a current-year CAD/CAM/CAE value. These interrelated concepts are important in characterizing real-world issues present in all programs. Accurate characterization is necessary to develop an accurate BCA. For each flight system equipment item, a historical analogy is selected, and the model changes historical costs to the TGALS system conditions serially by dimension. For example, the glider was compared to the Lockheed U-2. The U-2 historical costs, physical description, and technological conditions were calibrated into ICAM. Subsequently ICAM changed the U-2 into a TGALS glider one step at a time through all technology and physical dimensions. These steps included demilitarization as well as modernization.

B. Organization Complexity One disadvantage of the TGALS concept is the relatively complex organizational structure required due to the diversity of flight system component types. The tow aircraft (an existing business jet or aerial tanker), glider, sustainer motor, and launch vehicle are designed and built by different companies, necessitating a sophisticated management, procurement, and logistics structure within the TGALS organization. The Aerospace cost modeling system increased TGALS integration and management costs to capture the more complex business and engineering environment as compared to a “flat” organization that builds and operates standard launch vehicles.

C. Market Size Forecast The microsatellite market forecast by SpaceWorks Enterprises, Inc. (SEI) in 20162 predicts an escalating demand for launch services for the smaller satellites that is driven by the commercial sector’s interest. In order to forecast market launch rates, annual smallsat launch rates were analyzed and annual growth rates measured (Figure 6). To account for multiple manifesting per launch, projected annual launch rates by payload capacity were projected. The TGALS study displayed the scalability of the launch concept for satellites up to 550kg.

6 American Institute of Aeronautics and Astronautics

Forecast of Launches for 5 to 50 kg Payloads

1,000 Exponential Forecast (optimistic)

Annual Sat Launches Forecasted Launches Manifesting = 2 to 5 per launch 5 kg (actual and forecasted) 15 20 100 25 Logrithmic forecast (conservative)

Launches 10

Annual Small Sat Launches (actual and forecasted)

1 2000 2005 2010 2015 2020 2025 2030 2035 2040 Year

Figure 6. Market Forecast

The launch forecast has variability resulting from a significant number of proposed new smallsat LV entrants, each with different manifesting strategies and payload capacities. Not all of these will be launched. However it should be noted that the forecast used is very conservative in the out years as compared to the historical trends. Even the early years (2017-2025) are conservative as they do not account for the payload backlogs resulting from the launch failure down times that occurred in 2014 and 2015.

IV. Evaluation Process

A. Program Assumptions Technical and physical equipment, ground infrastructure, and task descriptors are input into the modeling system to document the process and approach. This includes task descriptors and technological metrics to components as small as the tow cable reel mechanism and the cable within the tow aircraft. These inputs are too numerous to list in this paper. Some of the main cost and market drivers are listed below:

 Market assessments and financial returns include operations to 2040  Profit margins charged are reflective of marketplace competitiveness  New flight providers to Low Earth Orbit (LEO) = 11, survival rate for new providers = 70%  Each flight vehicle design approach is assumed successful  No disruption due to catastrophic failure is included  Hosted payload market is minimal after 2017 due to increase in small LV providers  Flight growth rate to 2030 = 19% compounded (The 2012-2015 smallsat growth rate was 27%).  Tax Rate = 35%  No insurance is included  Various payload capacities  Tow aircraft range from business jets to firefighting tankers  Tow aircraft rented on demand or fully owned

7 American Institute of Aeronautics and Astronautics

B. Financial Metrics Modeling of TGALS performance, physical characteristics, development costs, flight test costs, infrastructure costs, market size, market capture, resultant production quantities, operations costs, gross revenue, and profits are all necessary components for calculating financial evaluation metrics. Three commonly used financial metrics were used to evaluate TGALS as a viable business venture, i.e., ROI, NPV, and PBP:

 Return on Investment (ROI) ROI is a financial investment metric used to gauge the efficiency of alternative investments in selecting the best investment. ROI is a measure of financial return on the investment cost. ROI is popular due to its simplicity of calculation and wide applicability. A weakness of ROI is that it does not consider the amount of time required to achieve the return. ROI = (Gain from Investment – Cost of Investment) / (Cost of Investment)

The ROI calculated for the TGALS program as envisioned is conservatively 8.3% based upon launch rates of 8 to 22 launches a year. The profit margin on the launches is based upon price reductions from competitive launch rates. The predicted ROI appears realistic because (1) conservative inputs were used in the analysis in areas where uncertainty exists and (2) a large number of elements in the WBS was input to assure completeness of the cost estimate.

 Payback Period (PBP) PBP is often used to supplement ROI as it is simple to calculate and accounts for the time required after investment to recoup the investment. As such, PBP does not indicate profitability of a project, although it can be cautiously assumed on some projects that a rapid PBP is usually more profitable that a longer PBP. Time spreads of TGALS estimated costs and revenue indicate a reasonable PBP of 6 to 8 years. This PBP value is short for launch vehicles due to the large use of existing flight components and simple infrastructure composing the TGALS system.

C. ROI Variable Uncertainty Palisade @RISK, a commercial Excel add-in package, was used to quantify the impact of input uncertainty on the model’s output. Palisade uses the Monte Carlo method of statistical distributions to simulate the BCA model results when operated thousands of times within the ranges specified for selected input variables possessing substantial uncertainty. This method has been used extensively in the aerospace cost estimating community since 1965. This section will discuss the inputs, outputs, and sensitivity of this analysis.

 Resulting Probability Distribution and Sensitivity Simulation of the TGALS program 5,000 times was performed with the six variables having substantial uncertainty varied within specified ranges, to create Figure 7. The six variables were LV recurring cost, profit margin, annual launch rate, glider maintenance cost, payload capacity utilization, and estimate reserves. Using conservative data interpretations and input ranges and a comprehensive WBS, the resulting ROI financial metric was the lowest, at -3%, when all variables were simultaneously pessimistic. This is not expected to occur. Similarly, all optimistic variables happening simultaneously yielded a ROI of 24%. A more useful statistic is an expected ROI range from 3% to 14%, which includes 90% of all expected input value combinations. The most realistic occurrences of the variables provides a mean ROI expected value of 8%. Considering that the input values and analysis are conservative, the 8% ROI is favorable.

8 American Institute of Aeronautics and Astronautics

Frequency Distribution Distribution

Baseline ROI, %

Figure 7. ROI Probability Distribution

Another way of observing the ROI output is by using the cumulative probability distribution shown in Figure 8. As an example, the reader can move horizontally to the right from 0.5, which represents 50%, to the ‘s’ curve and then proceed down to 8%. This shows that there is a 50% probability of 8% ROI. Reading horizontally from 0.8 (80%) to

the curve and down shows that 80% of all occurrences of ROI simulation are below 12%.

Probability

Cumulative

Baseline ROI, %

Figure 8. ROI Cumulative Probability

As shown in Figure 9, several input variables dramatically drive the ROI. The variables selected for drivers of uncertainty include LV recurring build cost, profit margin chargeable in a competitive market, annual launch rate, glider maintenance cost, amount of total payload capacity utilized by multiple manifests, and estimate reserves. Estimate reserves are extra costs applied to the cost estimate for unknowns. The variables are listed in Figure 9 in

9 American Institute of Aeronautics and Astronautics order of importance driving uncertainty. In this case the recurring LV cost and the chargeable profit rate each play significant roles.

Baseline ROI, %

Figure 9. Inputs Driving ROI Uncertainty

V. Summary

A. Observations Several notable observations were derived from the BCA analysis: 1. The effect of a long and expensive flight test program can be devastating to financial success metrics. A long flight test program will add expense and push the revenue stream further out, delaying cash flow. On the other hand reduced testing increases the chance of a catastrophic failure, which would drive cost up even faster and delay financial returns even higher should it happen. The effects of a catastrophic failure on the TGALS program have not been modeled. A human safety event in TGALS is unlikely as the pilots of the tow aircraft are operating mature commercial aircraft and are separated from new system components by the nature of the system architecture. 2. Small LV market predictions are difficult in the small satellite market as multiple manifesting is a large uncertainty factor. Also large numbers of small satellites will be launched on large LVs as dispensers improve. Due to multiple manifesting, a LV failure with large numbers of small satellites can vary the total market by as much as 30% in a single year. For example, the actual launch rate in 2015 was 30% less than the SpaceWorks average annual launch forecast for 20152. The overprediction by SpaceWorks was caused by the Falcon 9 failure in June 20152 and the Super Strypi failure in November of 20152. The Super Strypi launch failure alone was responsible for the loss of 51 small satellites along with the primary payload2. 3. Multiple configurations of flight systems with a full range of payload sizes were studied. A problem was encountered with predicting capture rates when the larger payload TGALS configuration competed with the smaller payload configuration. 4. When focusing on total cost, a design change of $25 million does not significantly change total investment cost in the $500 to $1B program class. Yet it has a much greater impact on investment statistics such as ROI. The $25M

10 American Institute of Aeronautics and Astronautics may be an increase of 5% to 10% of development cost, but from the profitability standpoint it can push the break- even point and profits out as much as a year.

B. Conclusions Based upon extensive modeling and cost trade studies, a successful solution to the TGALS business case was developed. The is composed of mostly low risk components compared to typical new launcher development programs and is sufficiently robust in its components that alternatives exist that greatly enhance the business case viability. One major example of this is the rental of tow aircraft made possible by the minimum number of aircraft modifications that are necessary. Importantly, because the BCA model was comprehensive and detailed in description of design characteristics, programmatic, and market assumptions, all decisions and alternatives could be explored with the focus on financial returns for a commercial investor. A summary of the TGALS low-cost characteristics follows:

• No new infrastructure costs • No dependence on critical ground based launch facilities • Relatively inexpensive to build a glider vs. modifying a conventional airplane • Glider has low maintenance costs compared to a conventional airplane • Existing business jets and tanker aircraft can be used as tow vehicles with minimal modifications

Although the trade space explored in the TGALS BCA was too large to discuss here in entirety, Figure 10 provides an example of results from one trade study under defined successful conditions. Based upon the number of launches ranging from 5 to 25, this TGALS payload capacity became highly competitive in price with a 30-60% margin above 10 launches per year. The ROI gained attractiveness at 10 launches per year with a 60% margin. These margins appear feasible as a function of the physics advantage of the towed glider coupled with a robust and flexible concept composed of relatively low risk technologies.

Price per Flight, FY16$ Return on Investment (ROI) 20 35% Average Competition Price Profit Margin % 90% Point 18 30%

Profit 25% 16 Margin % 60% 90% 20% 14

ROI ROI % 15% 12 60% 30%

10% Price Price per Flight $mFY16 10 30% 5%

8 0% 5 10 15 20 25 30 5 10 15 20 25 30 Annual Launches Annual Launches

Figure 10. Price per Flight and ROI versus Annual Launches

A healthy and robust launch rate is necessary for a successful business case for launch vehicles. The baseline analysis of 60% profit margin yields an 8% ROI on 12 launches. But as the right side of Figure 10 shows, a capture rate of 20 launches a year (+67%) doubles the ROI to 16%. A launch capture rate of 20 per year, although historic, is not unreasonable for a system with mostly reusable components such as TGALS. This is true especially considering the conservative nature of the BCA performed. A summary of the conservative BCA elements follows: 1. The WBS was very detailed in all categories from development and production costs, operations costs and infrastructure.

11 American Institute of Aeronautics and Astronautics 2. The model accounts for changing technology, advanced design processes, and industry skill levels. 3. The model inputs passed a grueling ‘vetting’ by Aerospace Commercial Launch and Systems Engineering cost experts midyear 2017 to assure a lack of ‘wishing and hoping’ so commonly characteristic in BCAs. 4. Estimate reserves were added to account for any ‘I forgots’ or ‘unknown unknowns’ in addition to the uncertainty simulation. 5. The launch forecasts do not account for the increased ease of using new launch adapters or future smallsat launch surges due to delays in manifesting caused by large LVs. 6. Analysis included changing market conditions and improvements in the positions of competitors as opposed to the market remaining constant in manifesting, efficiency, and competitive pricing strategies.

In summary, the model indicates that TGALS variants present viable positive business cases.

Acknowledgments The authors are grateful to Angela M. Monheim, Jacobs Technology, Inc.; Kenneth J. Szalai, Aerospace Services International LLC; Charles Rogers, Armstrong Flight Research Center, and Glenn W. Law, The Aerospace Corporation, for their valuable guidance and support.

References 1Webb, D. W., “The Clash of Technology, Time, Performance and Cost,” Space Systems Cost Analysis Group (SSCAG), London, September 19, 2006. 2SpaceWorks Enterprises, Inc. (SEI), “2016 Nano/Microsatellite Market Forecast,” pp. 3 and 6. http://spaceworksforecast.com/docs/SpaceWorks_Nano_Microsatellite_Market_Forecast_2016.pdf

12 American Institute of Aeronautics and Astronautics