In aviation terms, Pratt & Whitney (P&W) has performed a steep ascent and is now preparing to settle in for what it hopes will be a long and turbulence-free cruise — but a cruise whose early stages will see the company perform its fastest-ever increase in production of aero engines.
Speaking to reporters on April 2, P&W president Paul Adams said the OEM had “been in a very significant reinvestment phase for the past several years”. However, Adams added the company had negotiated this critical period successfully and was now at a point where it could expect to begin enjoying the fruits of its hard labour and intensive spending. “Our certification programmes are very much on track and are performing the way we expected,” he said.
The most important of these certification programmes for P&W’s future are those for its PurePower PW1000G family of geared-turbofan (GTF) engines for commercial jets. Five aircraft manufacturers have chosen the PW1000G for new airliner families – Airbus for its A320neo family, Bombardier for the CSeries, Embraer for the E2 family, Irkut for its MC-21 and Mitsubishi Aircraft for its forthcoming regional jet.
It is already clear that at least two and possibly all five of these manufacturers will see high sales of their new aircraft. So it is highly important for P&W that its certification programmes for the five PW1000G families go well. Two — the PW1500G for the CSeries and the PW1100G for the A320neo — have already been certified and both models should enter commercial service by 2016. But much lies ahead for the PW1000G, even in the remaining months of this year.
“This is an incredible time for us,” said Greg Gernhardt, president of P&W Commercial Engines. “This year our PW1200G engine will power the Mitsubishi Regional Jet on its first flight, our PW1400G engine for Irkut will complete testing and receive certification, and our PW1900G engine for Embraer’s E190-E2 and E195-E2 will initiate its flight-test programme.”
By February, P&W had accumulated orders and options for more than 6,300 PW1000G engines. The majority of the sales are accounted for by the PW1100G which holds an approximate 45 per cent share of the A320neo-family orders against CFM International’s LEAP-1A — though customers for about 40 per cent of all A320neo-family orders haven’t yet specified their engine choice.
On April 2, when confirming the long-held industry suspicion that the first PW1000G model to enter commercial service (in the second half of 2015) would be the PW1100G on the A320neo rather than the PW1500G on the Bombardier CSeries, Adams said he was unconcerned by any perceived market-share difference at this point.
His professed nonchalance on the subject is understandable given that the LEAP-1A engine only entered flight tests on the A320neo on May 19. The 1A follows in the footsteps of the 1B which took its first flight on a 747 testbed on April 29 and the 1C which entered the flight test programme in October 2014.
“I’d dispute that 10 per cent share differential on the A320neo — those numbers are manipulated,” said Adams. That said, “I’m not wound-up about market share per se. I think we’re in a great position. We are able to say to customers: ‘Here’s what we can do — what can the other guy show?’ This is a long-term game and I don’t get wound-up about share tomorrow. These are long-term decisions for airlines and they will go both ways.”
That includes airlines switching from being CFM International customers to P&W customers. Adams noted that, as of April, seven airlines which had ordered the A320neo family and previously had been CFM56 customers for their single-aisle aircraft had switched their allegiance to the PW1100G.
Meeting performance promises
One reason for Adams’ confidence in his product is that, after more than 3,000 hours of flight-testing and 16,000 hours of total testing in which a fleet of more than 50 development, test and compliance engines have operated the equivalent of 31,000 flight cycles, P&W has been able to confirm to its first two airframe customers the performance guarantees it gave for the PW1000G. With the CSeries certification planned for this year, Bombardier is maintaining its boast that the all-new aircraft will be at least 20 per cent more fuel-efficient than current competitors.
For the PW1100G’s initial entry into service, P&W has guaranteed Airbus that on a typical two-hour sector the A320neo will offer 16 per cent lower fuel burn than an A320 fitted with V2500 Select engines (the latest build standard for the V2500). It will also produce less than half the emissions produced by the V2500 Select-equipped A320 and — astonishingly — its engine noise levels will be at least 75 per cent lower than those produced by aircraft powered by current comparable engines, according to Adams.
When the PW1100G enters service, Adams thinks A320neo customers which haven’t yet made their engine choice will be forced to take notice of what the PW1100G is offering. “The good news is that the [PW1000G] technology is performing at a very high level of performance reliability,” said Adams.
But the good news doesn’t end there. P&W has already agreed with Airbus that, when the PW1100G enters service — a milestone expected in October or November when Qatar Airways takes delivery of the first A320neo — it will guarantee to deliver by 2019 a package of PW1100G improvements that will reduce each A320neo’s fuel burn by another two per cent.
So confident is P&W in the GTF’s performance that in March it raised the prices of all its PW1000G engine models. Although the company hasn’t said how large the price increases were, Paul Finklestein, P&W’s VP marketing for commercial engines, commented to ATE&M: “The message was clear to the marketplace — we expect the value proposition to be excellent to the customer.”
Higher thrust, weight on target
Also, not only has P&W certified the PW1100G for its originally planned maximum take-off thrust of 33,000lb for the A321neo, but the company has been able to certificate the engine for 35,000lb of maximum thrust. (The engines powering A319neos and A320neos will have lower maximum-thrust levels governed by data plugs in their engine control modules, but all PW1100Gs will have identical hardware.)
Adams assured ATE&M that the original 33,000lb certification target would be quite enough thrust to power the recently announced 97-tonne A321LR transatlantic-range version of the A321neo, but added that the 35,000lb certification will be important for operators whose networks include routes from airports in very hot climates or at high altitudes. Since several A320-family customers in South America, China, Central Asia and the Middle East serve such airports, they may be prime targets for the engine option.
Danny Di Perna, P&W’s senior vice-president of manufacturing and operations, said the OEM is also “on specification” for its guarantees to Airbus and Bombardier regarding the weight of the PW1100G and PW1500G engines powering their respective aircraft families, and also for the weight of the PW1100G’s fan drive gear system (FDGS).
The FDGS is the near-revolutionary, compact assembly of star, planetary and ring gears which are made of decidedly non-advanced gear steel, but which together create the GTF’s compelling performance edge. They do so by translating the 10,000rpm rotation rate of the PW1100G’s low-pressure spool into a rotation rate of less than 3,500rpm for the shaft which drives the engine’s fan, increasing its propulsive efficiency. The production FDGS for the A320neo’s PW1100G weighs a mere 113kg, according to Di Perna — a 90kg reduction from the 180kg-plus weight of pre-production test versions.
Having a promising engine to offer customers is all very well, but P&W faces a daunting task in terms of increasing PW1000G production from virtually zero to as many as 1,800 engines a year by 2020. By 2017, PW1000Gs will be flying with 39 customers, only 23 of which are Pratt & Whitney customers now. PW1100G production has to hit approximately 550 engines a year within four years, according to Di Perna – noting that it took 25 years for V2500 production to reach the current rate of approximately 500 engines annually.
P&W has developed two very different ways of coping with this enormous challenge. One has been to identify from the start that PW1000G production would run on multiple assembly lines located in different parts of the world. PW1100G assembly will be performed on three lines: at Middletown in Connecticut, P&W’s largest production facility; at its new West Palm Beach line; and at a line run by MTU Aero Engines in Munich.
Meanwhile, PW1500G production will take place on a line P&W has established at Mirabel Airport near Montreal in Canada, located right next to the CSeries final assembly line. Production of the PW1700G and the PW1900G for the E2 family will take place at Middletown and West Palm Beach, but P&W is seriously considering locating a new PW1700G (and possibly PW1900G) line near Embraer’s assembly line at São José dos Campos in Brazil. PW1200G assembly, meanwhile, will likely take place in Japan, on a line run by P&W’s PW1000G programme partner Japanese Aero Engines Corporation.
P&W’s second solution to the production ramp-up problem is to create advanced, automated moving assembly lines for the engine’s manufacture. Its test-bed for the technique has been the PW1500G line at Mirabel Airport, the first PW1000G line the company established. So successful did its initial experience with the moving line prove that P&W has made its PW1000G lines at Middletown and West Palm Beach moving lines as well.
Each of these loop-shaped lines has what Di Perna calls “zero-gravity hoists” which hold each PW1000G horizontally — rather than vertically, the traditional way of building engines — to make it easier for assembly mechanics to gain access to every part of the engine. Assembly starts with the core modules and then continues with the installation of the FDGS, fan and fan casing, exhaust nozzle, accessories and external fittings.
Each zero-gravity hoist (there are 20 of these on the Middletown line) is counter-balanced so that with just a touch of a hand a technician can position an engine — by rotating, raising or lowering it — to the position he or she finds most comfortable and convenient for the assembly task to be performed.
At all times each mechanic is provided with the relevant parts for the assembly step upon which he or she is working, the parts being brought right next to the very slow moving line. The line has barcode badge readers at every assembly station and each engine is identified with a barcode label so that, when an engine reaches a given assembly station, the barcode reader automatically moves the engine along to the next station and informs the engine work database. The database software then creates notifications for the mechanics to be provided with exactly the right types and numbers of parts for the next assembly step at the new station.
Di Perna said that with the moving line each PW1100G will take less than five working days to assemble. The line has eight test stands for production pass-off. Each moving line also has separate loops so that engines can be separated into two groups: those which have yet to undergo testing in the two production test-cells located adjacent to the line; and those which have already been tested. Each PW1100G only spends about two hours in the cell for production testing, according to Di Perna.
So well is the new moving-line system working that Di Perna estimated it will provide “20 to 30 per cent more productivity” than a traditional assembly-line arrangement. Assembly of the first engine on the Middletown moving line took only half the time of previous engine builds.
“We will only get better,” said Di Perna, acknowledging that the company doesn’t know “everything” about the moving-line system yet and that more experience of operating it will bring greater efficiency. In fact, he estimated that even though P&W’s Middletown and West Palm Beach lines — which are identical in design — will each be producing 250 to 270 PW1000Gs a year in short order, they are each capable of producing more than 500 annually.
In the 1960s and 1970s P&W had dominated the entire commercial-jet engine market, but in the late 2000s it had fallen to a distinct second-best in the single-aisle market and a distant third place in the widebody market. Its GTF-fuelled resurrection from fading giant has profound implications for its future place in the aftermarket.
So much is this the case that Adams highlighted aftermarket services as a fundamentally important plank of P&W’s future business. The company’s traditional aftermarket business model has mainly been transactional — selling spare parts and spare engines to customers rather than helping to manage their fleets on a rate-per-flight-hour basis. Only 40 per cent of P&W’s PW4000 aftermarket business and 60 per cent of its V2500 aftermarket business lies in such “fleet management programmes”, according to Adams.
However, P&W is determined to move to this sort of aftermarket model for all its future commercial engines, in what Adams described as “a very significant change” for the company. Its aim from now on, he said, is to create “long-term contracts aligned with airlines’ goals, and make sure our goals and airlines’ goals are the same — to make sure their engines are reliable and stay on-wing as long as possible”.
About 75 per cent of the aftermarket in the GP7200 joint-venture engine that P&W and GE Aviation make for the A380 is composed of such agreements and P&W has agreed fleet management programmes for 80 per cent of all PW1000Gs ordered to date. In fact, P&W’s backlog of aftermarket service programmes has increased 50 per cent since 2010 to $57bn, its commercial engines accounting for more than two-thirds of the total.
According to Adams, keeping engines on-wing longer, doing as much on-wing maintenance as possible and achieving lower engine lifecycle costs for its customers through fleet-management programmes will increase P&W’s own profitability, by encouraging more customers to sign up for long-term fleet management agreements for more engines. Critical to P&W’s entire new service-based aftermarket business model is big data — or, as the company prefers to call it, “predictive analytics”.
Matthew Bromberg, president of P&W’s aftermarket business, said the company has been collecting engine health monitoring (EHM) data from its engines for 20 years, but only now has it reached “an inflection point” where its business requires P&W to use such data in powerfully analytical new ways. Several major technological advances have also been necessary for the company to be in a position where it can do so.
Today, P&W collects data covering 100 performance parameters per second from its modern commercial engine models — the GP7200, late-model PW4000s and V2500. However, in the PW1000G and future GTF engines the company will be able to collect data on 5,000 performance parameters per second.
Collecting this river of performance data from every engine in the fleet will create a staggering ground-storage requirement for as much as 12 petabytes — 12 million gigabytes — of engine performance data. This is equivalent to six times the entire amount of data currently stored in all US research institutions, Bromberg calculates.
But exponentially greater data capture is only one factor among several that together ensure P&W can move to a service-driven aftermarket model. Unless P&W can crunch floods of EHM data in ways which can predict when unscheduled-maintenance events are likely to happen and allow the OEM and its customers to stop them from occurring in the first place, its data-driven efforts to monitor engine health will mean little.
Larry Volz, P&W’s chief information officer, said advances in database and storage activity — particularly its move to direct data storage in stable solid-state drives and high-performance drives — represent one factor helping the engine manufacturer make its predictive maintenance vision a reality. Another is P&W’s move away from using standard database structures towards massive parallel computing, allowing it to stream and analyse data at the rate of 10 gigabytes a second.
A third important element is the maturation of software which can aggregate and analyse structured and unstructured “asynchronous” data together, said Volz. Such software can take performance-parameter data feeds from engines and combine them with PDF files of engine incident reports, for example, or field representatives’ reports, and from this asynchronous data develop a very accurate idea of what is most likely to go wrong with a given in-service engine and when it is most likely to happen.
Volz also highlighted two other developments critical to P&W in advancing its predictive-maintenance agenda. One is its identification of IBM as a leader in the field of big-data analytics, IBM having invested $24bn in this area over the past four years. P&W has signed a partnership agreement with IBM to act as “an accelerator” in the expansion of its big-data expertise from the engine-design and airflow-modelling fields in which it has long held competency to new areas such as predictive maintenance.
The other is P&W’s current move of all its data into “the private cloud and virtual private cloud offerings”. Not only will this secure P&W’s intellectual property — “in the public cloud you have too many neighbours” — but Volz said private and virtual private networking arrangements will gave it “flexibility and speed”, making it “able to flex up and down in terms of [data] capacity”.
The power of predictive maintenance
Using historic data, P&W experimentally modelled its capability to predict inflight engine shutdowns for the in-service fleet of PW4000 engines. The project was one of 14 developed under the management of Lynn Fraga, P&W’s analytics manager for engine services and leader of a 30-member team spearheading the company’s predictive-maintenance effort. Although the computer model generated a few “false positives” (predicting events which didn’t actually happen), it accurately predicted some 90 per cent of PW4000 in-flight shutdowns before they occurred, according to Bromberg.
Any given P&W commercial turbofan engine may only have an unscheduled removal once in 100 years and its entire fleet of the engine model may only have such an event once a year. However, Bromberg said: “I can easily see a 50 per cent reduction in unplanned engine events” as a result of P&W applying its predictive-maintenance modelling to any fleet of its in-service engines.
The company expects to finish developing an operational system by the end of this year and to test it in 2016, after which P&W will begin offering the service to airline customers. According to Bromberg, the result could be that, where today it takes an average of 10 people at a given airline to manage a single one of its engines fully, with predictive maintenance (and a P&W fleet management plan) it may soon take only a single airline employee to do so.
If Bromberg’s predictions prove correct the PW1000G family of GTF engines could prove to be revolutionary in the maintenance hangar as well as on wing. It’s no wonder that its entry into service is so hotly anticipated.