USAF Tankers: Critical Assumptions for Comparing Competitive Dual Procurement with Sole Source Award The Congress has expressed interest in better understanding the costs associated with competitive dual procurement of tankers compared to current plans to award a sole source contract for KC-X followed by KC-Y. There has been public dialogue advocating significant cost savings associated with programs acquired using continuous competition compared to significant cost growth associated with sole source procurement. The DoD is on record stating that dual procurement will cost $7 - $14B more than the proposed program of record which would procure a single platform. In that regard, it is important to ensure that any cost analysis appropriately reflects all costs associated with the anticipated program of record as well as efficiencies that could be achieved by functionally accelerating KC-X and KC-Y execution under a "competitive dual procurement" approach. The attached paper is provided for your consideration. The summary table articulates the major assumptions that must be considered when comparing a sole-source for KC-X and KC-Y to a dual award with continuous competition. The Appendix provides more detail on each of the assumptions summarized in the table. While the cash flow associated with a competitive dual procurement approach is accelerated when compared to the program of record, the higher level of investment delivers urgently needed capability to the warfighter more quickly than the current plan. Further, the analysis and assumptions presented in this paper indicate that a competitive dual procurement approach would yield significant total acquisition and cost savings/avoidance when compared with the anticipated program of record.
USAF Tankers: Critical Assumptions for Comparing Competitive Dual Procurement with Sole Source Award Tanker fleet cost-to-go over acquisition timeline 1 Number of tankers to be replaced Replacement rate KC-135 O&S Cost Flying hour assumptions System Design and Development (SDD) Procurement cost growth/reduction Production rate Potential efficiencies Logistics & Training Tanker Fleet Cost To generate a meaningful comparison, the 30 year cost-to-go of the 4 tanker fleet required in the sole source award (KC-135, KC-X and Y, and KC-10) and the 4 tanker fleet in the dual award (KC-135, KC-45, KC-767, and KC-10) must be calculated. The USAF should replace a total of 415 KC-135R equivalents. The USAF calculated the capability of the KC-45 and KC-767 in terms of KC-135R equivalents using a sophisticated methodology during proposal evaluation. The USAF should use the calculated KC-135/KC-45 replacement rate of 1.9:1 and the KC-135/KC-767 replacement rate of 1.79:1. A KC-135 costs more to fly per hour than the KC-45 or KC-767; the legacy fleet also faces a doubling of depot costs. Credit should be given for accelerated retirement of the aging KC-135 fleet. Overall support costs of the tanker fleet will be reduced. The KC-X RFP states new tankers will each fly 750 hours per year compared to the KC- 135 s 500 hours per year, even though the new tankers offer significantly more capability. For a realistic comparison, increased flying hours for the new tankers should be matched by decreased KC-135 hours or all tankers should fly the same number of hours per year. Development and Production The recent program of record requires an SDD 2 for KC-X, followed by a second SDD for KC-Y, and then a third SDD for KC-Z. Three separate cycles of SDD is costly. A dual award is analogous to simultaneously conducting SDD for KC-X and KC-Y, and provides opportunities for concurrent risk mitigation and testing across both KC-45 and KC-767 aircraft, including a reduction in the required quantity of SDD aircraft. The planned procurement will likely yield a sole-source environment for KC-Y and KC-Z phases which encompass more than 300 aircraft since there is no business case for the losing KC-X offeror to remain in the market. Sole source programs historically encounter significant cost growth. Dual award could yield continuous competition for all 500+ aircraft, resulting in significant savings. A parametric analysis of the implications should be presented. Analysis should use the planned production rate for KC-X and KC-Y (about 15 per year for ~$3B) for sole source and a rate of 24 per year for ~$5B) for continuous competition dual award. An increased production rate also accelerates program, resulting in significant escalation savings. The dual source award will feature higher production rates. Since the two manufacturers use many common suppliers, economic order quantities could generate savings. In addition, the government could specify commonality in some GFE to generate additional savings. Under dual award, the USAF establishes the same logistics and training infrastructure as the KC-X and KC-Y program of record, though over a shorter period of time. Dual award retires the higher cost KC-135 logistics/training system earlier, and also allows efficiency through commonality across KC-45 and KC-767 that was not achievable due to the serial procurement of KC-X and KC-Y. 1 This period is recommended because completing a KC-X and KC-Y program at the planned rate of 15 per year will take about 30 years from now considering development timelines and the time required to conduct a second competition. For an accurate comparison, both programs should be assessed over the same time interval. This is distinguished from an RFP MPLCC assumption of 25 or 40 years 2 SDD is now called EMD by DoD 5000.02 2
APPENDIX Tanker Fleet Cost-To-Go Over Acquisition Timeline The new tankers will replace the KC-135s, which cost more per hour to fly than the newer tankers and are experiencing a significant increase in depot support costs. Accordingly, the faster the KC-135s are retired, the less the USAF needs to spend on KC-135 support and the more that is available to spend on new tankers. In order to make an accurate comparison, the cost of the new tankers cannot be considered in isolation, but must be examined within the context of overall tanker fleet cost. The government s current acquisition strategy is to first procure the KC-X followed by the KC-Y about 8 years later. The dual award would accelerate this plan. For an apples to apples comparison, the cost-to-go of the entire tanker fleet for 30 years should be estimated. 3 Number of Tankers To Be Replaced The USAF is currently in the process of retiring 134 KC-135Es and all will be retired by the end of 2009. Accordingly the USAF s KC-135 fleet will be composed of 415 R/T 4 models. Replacement Rate Since KC-135 fleet O&S currently runs at ~$2.5B per year, the rate at which these aircraft can be retired can have a significant impact on overall tanker fleet costs. The USAF has not been specific on what the rate would be. During the KC-X tanker competition, USAF public materials postulated a 1:1 replacement rate. However in 2004, the USAF submitted a different future force structure plan to Congress. To develop that plan, the USAF, which at the time was planning to procure KC-767s, planned to retire 1.5 KC-135s for each KC-767 that entered service. Maintaining equal capability with the current KC-135 force offers a useful guide, particularly when combined with the USAF s recent analysis of the capabilities of the two new tankers. During the proposal evaluation for KC-X in 2007/8, the USAF conducted a highly sophisticated analysis called the Integrated Fleet Aerial Refueling Assessment (IFARA) to calculate the capability of each competing offering. This analysis factored in wartime demand, basing constraints, deployment requirements, defense of the homeland, mission capable rates, the impact of tankers refueling tankers, training, and depot rates to estimate the capability of each new tanker in terms of how many KC-135Rs in the total inventory it could replace. 5 The KC-45 was rated at 1.90 KC-135R equivalents; the KC- 767 at 1.79 KC-135R equivalents. The calculation thus indicated that if the KC-45 was procured, 1.9 KC-135Rs could be retired while still maintaining equal refueling capability. 3 This period is recommended because completing a KC-X and KC-Y program at the planned rate of 15 per year will take about 30 years from now considering development timelines and the time required to conduct a second competition. For an accurate comparison, both programs should be assessed over the same time interval. This is distinguished from an RFP MPLCC assumption of 25 or 40 years 4 A KC-135T is an R fitted with a refueling receptacle. 5 Typically, KC-135R equivalents refer to possessed aircraft (and thus do not count the aircraft in depot). In the USAF calculation during proposal evaluation, the computed Fleet Effectiveness Value or R equivalency was calculated for the total KC-135R inventory (thus factoring in training and depot aircraft). 3
Assuming an equal mix of 180 KC-45s (1.9 R equiv) and 180 KC-767s (1.79 R equiv), that force would generate a potential equivalent of 664.2 R equivalents. 6 This fleet would thus generate more capability than the current KC-135 fleet and thus could even allow retirement of some of the KC-10s. The dual procurement option, which features higher production rates, would allow the USAF to retire the KC-135s earlier than the sole-source option. When all the KC-135s are retired, that aircraft s extensive logistics infrastructure would also be eliminated, thus saving substantial funds. KC-135 Operations and Support (O&S) Cost O&S cost for the KC-135R/T fleet can be derived from the Air Force Total Ownership Cost (AFTOC) data base. As shown in Figure 1, according to USAF projections, KC-135 depot costs, which comprise a portion of overall O&S, are slated to rise substantially starting in the 2018 timeframe and will stay at these higher levels for approximately two decades. Much of the additional work involves re-skinning portions of the aircraft. Accordingly, options that include a large KC-135 force over this period will require that more aircraft be overhauled and will thus cost more. KC-135 Maintenance Hours Per Aircraft 70000 Actual Projected HOURS 60000 50000 40000 30000 Corrosion preventative measures have moved structural replacements to the right approx. 10 years 15 Year & 30 Year Periodic Work 400 300 Aircraft 200 20000 10000 100 0 FY 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 General Field Tasks Hvy Struct O&A 1 Time Struct Rewire I Rewire II/III Rewire IV/V/VI Concurrent Mods CPC Horiz. STA Terminal Fittings Aft Body Skins Top Coat Horiz STA Upper Skin Upper Wing Skins Crown Skins (Corrosion) Center Wing Upper Skins Outboard Wing Unknown KCR Engine Strut Flight Director Flight Controls (MISTR) Concurrent Mods UDLMs Supporting Studies KC-135 Economic Service Life Study Tanker Requirements Study 2005 CRS Reports for Congress: RAND Common Replacement Asset Study Air Force Fleet Viability Board 2005 Air Force Aerial Refueling, 19 Sep 2005 GAO Tanker Study 2004 Lexington Institute s Modernizing the Aerial Air Force Aerial Refueling Methods, 5 Jun 2006 Refueling Fleet 2007 Air Force Aerial Refueling, 20 Mar 2007 DSB Task Force Report on Aerial Refueling Air Force Air Refueling: The KC-X Aircraft Requirements, May 2004 RAND s KC-10 AoA Alternatives Study 2008 Acquisition Program, 4 Apr 2008 2 Figure 1: USAF estimates of KC-135R maintenance hours O&S costs for the new tankers can be derived from the Air Force s life cycle cost estimates of the Northrop Grumman and Boeing proposals submitted in 2007. 6 (180 * 1.9) + (180 *1.79). To maximize the pressure of competition, the competitive dual award strategy outlined by Gansler envisions that whichever contractor offers the best value gets a larger portion of the total buy. So the total R equivalents offered could be either greater or less than the even split laid out above. 4
Flying hour assumptions In the RFP, the USAF planned to fly the new tankers at 750 hours per year for each of the 167 possessed aircraft (179*.935 =167). In contrast, the existing fleet flies 500 hours per year per possessed aircraft. To conduct an apples-to-apples comparison, analysts have two options: Maintain an equal number of fleet hours by flying the new tankers at the higher rate, but reducing the flying hours on the remaining KC-135s. This would have the benefit of conserving the aging KC-135 airframes and reducing overall O&S costs. Fly both the new tankers and the remaining KC-135s at the current rate of 500 hours per year per possessed aircraft. System Design and Development (SDD) The USAF already possesses estimates for the SDD (now called EMD by DoD 5000.02) of the KC-45 and KC-767, which should be sufficient to calculate the SDD cost of the dual award. In addition, there should be some potential cost savings from running the two development efforts concurrently: government staff and ranges, for example, could be used more efficiently. Another possibility would be to assess whether instead of testing 4 aircraft per competitor (for a total of 8), the development work could be done using just 6 aircraft (3 per competitor). For a comparable sole-source award for 360 aircraft, the government would conduct one SDD for KC- X and then, as laid out in USAF planning documents, a second sole-source award for KC-Y approximately 8 years later. Both the KC-X and the KC-Y would require an SDD and the life-cycle cost estimate should include both. Both the dual award and sole-source award would require two SDD contracts. In addition, the solesource award would also entail the expense of running a new competition. Data from the recent KC-X competition could be used for this purpose. Procurement Cost Growth/Reduction In 2006, Dr. Jacques Gansler, the former head of AT&L, published a study advocating the use of continuous competition on the tanker program. 7 He noted that the tanker program was an ideal case because the tankers were based on an already existing commercial aircraft. Accordingly, the SDD cost and the development of manufacturing facilities would be fairly low (compared to a dedicated military system like the C-17). And the benefits could be high. Gansler s data showed: 10 sole source aircraft evidenced 46% average cost growth 7 competitively procured aircraft averaged 16% cost reduction 20% average net savings dual sourcing missiles 1975-1995 13 studies on competitive sourcing 1964-1979 reported 32% average savings 7 Jacques S. Gansler and William Lucyshyn, Competition and the USAF s Tanker Replacement Program, University of Maryland, August 2006 5
Dual sourced missile systems had 75% less cost growth Critics noted that Gansler s evidence on aircraft cost savings involved commercial aircraft, not military systems, and that the majority of programs where continuous competition had been used were typically missiles and munitions, which have higher production rates and lower unit costs than tankers. Nonetheless, continuous competition did appear to offer the opportunity to harness the competitive posture between the two contractors to the benefit of the government. Three recent studies from the GAO and the RAND Corporation analyzed the cost growth in sole source DoD programs. All of these found that cost growth occurred in the majority of these programs. An earlier report by RAND in 2007 assessed the issue of whether weapon system cost growth was increasing, focusing on cost growth trends over the previous three decades. 8 This study relied on Selected Acquisition Reports data analysis, in this case through 2004. A general finding was that cost growth over the three decade period remained consistently high longer duration programs experienced higher cost growth and cost growth in developmental costs averaged 60%. Also worth noting, of the programs reviewed, a majority of aircraft development programs were found to have experienced average to higher-than-average cost growth. In a 2008 study, RAND identified specific causes of cost growth in its examination of 35 mature, but yet to be completed, major defense acquisition programs. 9 Based on analysis of data within SARs, RAND found that DoD historically underestimates the cost of new systems. For example, total cost growth on aircraft and helicopters averaged 74%. RAND concluded that total cost growth is due in large part to specific decisions, in turn accounting for over 2/3 of the growth. The study found that of these decisions, nearly 1/3 result from quantity and schedule changes. A 2009 GAO report on program cost growth examined 96 major defense programs in the 2008 portfolio looking at total program cost as well as unit acquisition cost. 10 It found in part a 42% increase in developmental costs for all programs while total acquisition costs expanded by 25%. Moreover, 42% of the programs looked at experienced unit cost increases of greater than 25%. The report also determined that older programs experienced higher levels of cost growth, with average growth ranging from 38% to 127%. Indications pointed to the conclusion that the longer a program runs, the higher the cost growth. In conducting a comparative cost estimate, the key issues are what assumptions to make regarding cost growth for a sole-source award and potential cost reductions for a dual source award. Gansler, for example, used a range of historical values to conduct his estimates and concluded that the worst case for continuous competition broke about even with the best case for sole source. For average savings, Gansler used an average cost growth of 46% for a sole source (based on prior examples) and an average cost reduction of 16% for the competitive dual sourcing case. In the case of 360 tankers at $175m per copy ($FY09) for a total program cost of $63B, the cost savings could be considerable. The sole source approach could increase costs by some $29B (46%) while the continuous competition could reduce the cost by $10B (16%). In this case the sole source 8 Obaid Younossi et al., Is Weapon System Cost Growth Increasing?: A Quantitative Assessment of Completed and Ongoing Programs, RAND Corporation, 2007. 9 Joseph G. Bolten et al., Sources of Weapon System Cost Growth: Analysis of 35 Major Defense Acquisition Programs, RAND Corporation, 2008. 10 United States Government Accountability Office, Defense Acquisitions: Assessments of Selected Weapon Programs, GAO-09-326SP, March 2009. 6
approach would cost $39B more than the dual award. Regardless, a comparative estimate should at the very least approach this issue using parametric analysis. Conversely, it is reasonable to assume that the winner-take-all strategy reflected in the program of record only supports a competitive market for the KC-X phase, after which the losing offeror would exit the market. Consequently, both the KC-Y and KC-Z phase of the program which encompass more than 300 aircraft would be awarded on a sole-source basis. Production Rates The dual award approach involves higher annual production rates 24 aircraft for ~$5B per year compared to the sole source award which averaged a production rate of about 15 (at ~$3B) per year. The dual award would cost more in the near-term, but, when combined with faster reductions in the KC-135 fleet, could offer considerable long-term savings. In addition, an increased production rate also accelerates the program by at least 7 years, resulting in significant escalation savings. Potential Efficiencies Production of at least 24 aircraft per year would yield economic order quantity savings with minimal tooling impact for both manufacturers as well as force furnished equipment. About 50% of Boeing and Northrop Grumman suppliers are the same companies, so the doubling of common suppliers should yield material savings. Maximizing the commonality of government furnished equipment and use of common suppliers/components for both types of aircraft could also save significant sums through economic order quantity savings. The government could require common self defense suites (for example, both the KC-45 and KC-767 already use the same infra-red countermeasures system), communications gear, engines (or common engine cores), and some elements of the refueling system. For example, the government could require that both aircraft be equipped with common refueling pods and potentially a common boom. Logistics and Training Under dual award, the USAF establishes the same logistics and training infrastructure as the sole source contract, though over a shorter period of time. The USAF purchases the same spares, trainers, and support resources under both approaches. For sole source, the USAF establishes the support infrastructure initially for KC-X followed by a second effort for KC-Y in later years. Dual source establishes support for both aircraft during the initial phase. Dual award offers the USAF a number of opportunities. First, the existing robust commercial support infrastructure provides opportunities to reduce support and training costs with shorter lead times and more available assets while maintaining the organic support posture. Second, by establishing the support capabilities for both aircraft earlier in the life cycle, cost of money impacts are avoided. Finally, dual award will retire the higher cost KC-135 fleet earlier, providing support cost savings over the life cycle, while improving capability and increasing readiness. 7