Early studies showed that four WR-21's in an IFEP configuration could propel a 30-40,000 tonnes CVF at a maximum speed of 30 knots.  However as CVF grew in size, speed inevitably dropped.  During 2002, with the BAE and Thales teams now considering CVF concepts of over 60,000 tonnes, the new Rolls-Royce Marine Trent MT30 became increasingly favoured over the WR21.  Because of it's significantly greater power output (36MW compared with 25MW), fewer GTA packages are required to achieve the required generating power. Despite being 45% more powerful than the WR21, the MT30 unit actually requires far less volume and weighs much less than the WR21 when the latter's complex intercooled and recuperating system is included.  Also, although no figures are published, it is likely that the MT30 costs less to procure, although over a 20+ year life cycle the higher fuel efficiency of the WR21 starts to tell in terms of saving on fuel cost and overall through life costs.
.

The Thales consortium quickly committed to IFEP for CVF and they brought Alstom on board as part of their team to provide expertise in electric propulsion and power distribution, bringing to bear their experience on Type 45 destroyers and elsewhere.  Thales' noted that the adoption of an IFEP system would allow the ship's electrical generators to be dispersed throughout the vessel giving improved survivability.  IFEP also opened the way to using podded drive propulsors ("pods"), a technology now common in modern cruise ships such as the 82,000 tonnes displacement (150,000 GRT) Queen Mary 2.

In conventional systems, electric motors are located inside the ship's hull.  With the new system, the motors are installed in pods fastened to the hull, which eliminates long shaft-lines.  Each pod includes a propeller. The pods can rotate a full 360°, so they do not only propel the ship - they handle manoeuvring as well..  One reason why pods were considered by both teams for the future carrier was to reduce the risk of delays and problems during build.  In a 'shafted' solution, the shaft itself, the bearings, thrust blocks and other components have to be put into the ship early on.  But instead, podded propulsion units can be added in just days at a later phase in the shipbuilding programme. Each pod can be fitted in five days, and this can be done at the end of the build.  Ship maintenance is easier, since the pods can be mounted and removed without moving the ship to a dry dock.  Using propulsion pods free up a large amount of space inside the hull, and the pods are quieter and generate less vibration.  Furthermore, podded propulsion also improves hydrodynamic efficiency by up to 10% (thus reducing fuel consumption) and manoeuvrability in confined waters and berthing. 

"Pods are a proven commercial solution and they offer us a lot of advantages, notably manoeuvrability and the flexibility they bring to the shipbuilding programme," said Thales' Robertson in 2002. "But there are some outstanding shock and signature issues."  BAE Systems took a similar view. "We are looking closely at podded propulsion because it holds a number of attractions," explains Chief Engineer Scott Whiteford. "But it is a technology not yet proven for the military environment. One option we are considering is a hybrid arrangement with a conventional centre shaftline and two podded drives." 

In its final AP2 proposals submitted in November 2002, Thales baselined a power train of 4 x 36MW Rolls-Royce MT30 gas turbine alternators (31MW electrical output) plus 5MW service diesel generators.  The ships generators were capable of 150MW in total, and were connected to the ship's systems by 2,000km of cable!  In the proposal, the CVF was fitted with a military version of the Mermaid pod unit similar to that installed in the Queen Mary 2.  The design had four 21.5MW pods totalling 85MW, each in a hydrodynamically optimised azimuthing body housing an electric motor in a "push" configuration.  The pod structure and prop-blade would be built by Rolls-Royce, and Alstom would supply the electrics.  Use of IFEP allowed the generator sets to be placed where ever the designers chose. 

"The reason we went for [pods] is they save us space within the ships, and for an aircraft carrier you can use them to manoeuvre more easily.  Each pod steers in 180 degrees and with the bow thrusters the vessel can self-berth." said Simon Knight of BMT Defence Services after the CVF design he had helped develop was selected in early 2003. "The principal worry with pod engines is the shock loading."  Pod engines have not shown a great resistance to shock. The effect has been to move the shaft out of alignment within the pod so that it comes into contact with other parts of the engine, causing the bearings to fail.  In the commercial market pod engines have used conventional roller bearings.  To improve the shock resistance the aircraft carrier design team went for sleeve bearings that offered greater and more constant contact within the bearing itself.

It is hard to translate the 85MW aggregate rating of the pods in to a traditional "shp" rating or an estimated speed based only upon the published data because this is their power consumption not output, combined with the hydrodynamic efficiency associated with pods (a reduction of 5-7% in drag is common), and also the slightly less than optimal hull form planned for CVF.  The final Thales proposal had a designed maximum speed of 28 knots on the equivalent to roughly 100,000 shp, by way of comparison, the old 50,000 tons (not tonnes) HMS Ark Royal [IV] had 113.4MW (152,500shp) on 4 shafts, for a maximum speed of 30.5kts

Meanwhile, the BAE Systems led CVF Team had slightly different ideas.  Rolls-Royce, who was an active part of BAE team and so were probably not giving Thales all its ideas, suggested 2 x 25MW WR21-ICR GTA's (21MW output), plus 2 or 3 MT30 GTA packages for high speed boost, powering a centreline shaft with wing pods.  With this configuration, the new carriers could normally cruise on the two very fuel efficient WR21's, but kick in powerful MT30's as "boost" when higher speeds were required. The partial commonality with the WR-21 powered T45's for spares and logistical support was also considered to be a significant advantage. 

Although the Thales CVF design was selected over the BAE design in January 2003 (and became termed the "Alpha" design concept), the Royal Navy quite liked the later's propulsion arrangements and apparently in February 2003 asked if the Thales/BMT CVF design could be similarly modified by the [now combined BAE/Thales] Aircraft Carrier Team, working with Rolls-Royce and Alstom.  However the idea seems to have been quickly discounted, and the configuration was actually adjusted to 3 rather than 4 x MT30 GTA's - the fourth gas-turbine had effectively been a 'hot spare' given the relatively low required speed of the ships - nice to have but difficult to financially justify.

The optimum location for the position of the main propulsion system was carefully examined in early CVF studies, with the need to maximize the hangar space below decks a major consideration.  The gas turbine generator units could be mounted in the superstructure, this would require a large island and reduce the flight deck area, but by avoiding volumous air intake/venting trunking to low machinery spaces will enable a larger and wider hanger.  The comparative advantages of the two layouts was extensively debated within the DPA and the two competing industrial teams, but operational analysis and aviation generation studies demonstrated that the extra flight deck space associated with a small island(s) would be more valuable than the extra hanger space, so traditional main hull located engine rooms were selected.  In the chosen Thales Alpha design, the two engine room units were widely separated, each one directly below an island to minimise the length of air downtakes and exhaust uptakes while offering good damage control.  This arrangement is possible thanks to the flexibility of IFEP and propulsion pods.  Battery's and several large diesel generators will provide emergency power if the prime movers fail for any reason.

In June 2003, the DPA asked the Aircraft Carrier Team to look at smaller CVF design in order to reduce costs, and changes to the engineering were considered.  One option was reverting to WR21 gas turbine's, but it was decided that dropping one of the MT30's was more cost-effective.   The use of pods and full or integrated electric propulsion, was also relatively expensive - a major concern for a project projected to be far over budget.  There were also continuing non-financial doubts about the suitability of pods for a such a large and important warship - including as described earlier.  Thus during the Autumn of 2003 the Aircraft Carrier Team, in conjunction with Rolls-Royce, carefully considered against the baseline the cost-benefits of several more conventional options that offered immediate savings in money, and possibly reduced weights and volume requirements:

1. 2 MT30 gas-turbine alternator packages driving 2 electric motors each powering a fixed conventional shaft;

2. 2 MT30 gas-turbine alternator packages driving 4 electric motors each powering a fixed conventional shaft; and

3. 2 MT30 gas-turbine mechanical drive packages with gearing directly driving 4 fixed conventional shafts. (A non-electric propulsion configuration)

24. [Baseline.  2 MT30 gas-turbine alternator packages plus 2 diesel generator sets driving 4 electric motors in propulsion pods.]

By October 2003 thinking favoured adopting option (A) for the smaller CVF design concepts being considered at the time.  A decisive factor was the intense pressure to cut project costs, (A) was the cheapest of the electric propulsion options considered in terms of initial build cost.   The solution was also very similar to the engineering architecture adopted for the Type 45 destroyers, but 50% more powerful - presumably sufficient to propel a 55,000 tonnes aircraft carrier at a maximum speed of at least 25 knots.

However it was recognised that with just two powerful gas turbines as the only prime movers, the CVF would be very restricted as to its one G-T cruising speed "window" for achieving maximum endurance, and the optimum cruising speed would not necessarily suit accompanying ships.  Also, aircraft carriers frequently change speed while operating aircraft and having to constantly double or half the power available is very uneconomic, another problem is that the loss of just one gas-turbine for any reason might prevent the carrier operating heavily loaded aircraft in low wind conditions.  It was found that considerable benefits would be accrued by adding several lower powered diesels generators to the remaining two gas turbines in an arrangement known as COmbined Diesel-Electric and Gas Turbine - CODLAG.  Fuel economy and increased range is one of several benefits.  The specific fuel oil consumption (SFOC) of a typical large gas turbine is much higher than that of a medium-speed diesel engine, which is very fuel efficient through a wide range of power output.  Gas Turbines have reasonably good specific fuel consumption (i.e. fuel efficiency) only when working near their maximum power output, consumption is particularly high at part load and typically exceeds 400 g/kWh as the load drops below 20%.  At maximum load, the consumption is still 207 g/kWh even for the modern and very efficient MT30.  By comparison with gas-turbines, the consumption of a large medium-speed diesel engine is extraordinarily frugal, below 175 g/kWh at high load.  Nor is the consumption of a diesel engine as load-dependent as that of the gas turbine; it increases only by about 30 g/kWh when the load is dropped to 20 %.  This large difference in SFOC between the gas turbine and the diesel engine gives rise to a typical characteristic of a CODLAG machinery - the fuel consumption is very low when the ship is operating on diesel engines alone, but increases rapidly when the gas turbine(s) are started.  But lower fuel consumption was not the only cost factor favouring the addition of diesel engines. Gas turbines also demand high-quality fuel such as marine gas oil (MGO), which is significantly more expensive than heavy fuel oil (HFO), normally used in diesel engines.  Indeed, the price of MGO can be twice as high as HFO.  Adding diesels to the machinery mix for CVF offered a compromise between low weight and space demand for high power that are best met by gas-turbines, and the good economy obtainable from a diesel-electric plant.