Turbo Tech: Internal vs. External Wastegates

by Khiem Dinh

Photos Courtesy of Bren Tuning and Killer B Motorsport


Before we get into the numbers and reasons why a dumped external wastegate makes more power than an internally wastegate turbo configuration, let's discuss why internally wastegated turbos are used by the OEMs. The first concern of the OEM is emissions; therefore all exhaust flow must flow through the catalytic converter. Can an external wastegate setup be configured so that the wastegate flow goes through the catalytic converter? The wastegate flow from an external gate most certainly can be plumbed back into the downpipe and into the catalytic converter. So, this brings us up to the next reason OEMs use internally wastegated turbos: simplicity.

The internally wastegated turbo comes to the OEM from the supplier as one part and therefore also one part number. The OEM hooks up oil lines, water lines, compressor and turbine inlet and outlet connections, and maybe a couple of vacuum lines or electrical connectors to this one turbocharger part. Let's say the OEM decided to go with an external wastegate configuration that was plumbed back into the downpipe. Now there are extra part numbers for the external wastegate and clamps, adding complication to supply base and logistics. On the design side, the OEM has to design more complex exhaust manifolds and downpipes with extra flanges and bellows joints. If the turbo is a twin-scroll setup, that’s double of everything. Then, the OEM has to do their own validation testing for durability and quality. Every extra weld and joint has to be validated for quality and testing durability. A 2000 hour durability cycle testing is common. As you can imagine, running an engine dyno cell for 2000 hours is not cheap (facilities, technicians, fuel, electrical power to the cell, etc). If the OEM gets the internally wastegated turbocharger all in one package, that greatly reduces costs and failure points.


This is the Honeywell (Garrett) turbo on the 2.3L Ecoboost engine on the Focus RS. It’s a twin-scroll with internal wastegate, and the compressor bypass valve is even integrated into the compressor housing. The turbo goes to Ford as one integrated and validated package and Ford bolts it to their engine.

The last area of concern for the OEM is the size of the package space required to fit the turbocharger. On turbocharged gasoline cars currently in production, the largest turbos are about GT25-28 size. That’s to say, they're not really big. BMW V8s use MGT2260s. The Focus RS uses a MGT2263. Single turbo cars are all in the 300-350hp range. Twin turbo cars are in the 600-650hp range, for the most part. So, the size of an external wastegate is quite large relative to these turbos. Finding space to fit external wastegates becomes a bigger problem in these often very cramped engine bays. Visualize trying to put a pair of external wastegates in the middle of the V of the German twin-turbo V8 engines.


This is the nearly stock BMW M6 engine in the Rahal Letterman Lanigan GTLM car with a pair of the stock Honeywell internally wastegated twin-scroll turbos. Notice how tightly everything is packaged around the turbos placed in the middle of the V. Imagine trying to package four external wastegates in there.

Okay, so we know why the OEMs go with internally wastegated turbos. But why does it give up power to externally wastegated configurations and how much power does it give up? The 'why' is easy to explain, the 'how' is much more complicated. The ‘why’ simply comes down to exhaust back pressure. Remember that the turbine wheel is an exhaust back pressure multiplier. Whatever back pressure the exhaust system creates gets multiplied by the turbine wheel. Bigger exhausts, test pipes, and high flow cats all allow the engine to make more power by reducing back pressure.


If you remember our article on turbo sizing, you’ll remember this data we generated by using the Borg Warner MatchBot program. The goal was to show how exhaust back pressure gets multiplied by the turbine wheel. In this example, at 8000rpm, a reduction of exhaust system back pressure of 2psi resulted in a reduction of pressure in the exhaust manifold by 4.4psi because the turbine was operating at a pressure ratio of 2.2.
Page 1 of 4 Next Page
Bookmark and Share
Mike Kojima
Mike Kojimalink
Wednesday, February 22, 2017 10:26 AM
Back in the day, I experimented with routing an external wastegate back into the downpipe vs dumping it to the ground on an SR20 and saw a 40 hp gain on a 400 hp engine at the same boost. It was a tial 44mm wastegate with a T3/T4 turbo.
Mike Kojima
Mike Kojimalink
Wednesday, February 22, 2017 10:35 AM
I am also of the opinion that a turbo like a Borg Warner EFR with an internal wastegate housing that is designed specifically for good wastegate discharge flow and a less turbulent entry into the exhaust stream will not be at too much of a disadvantage to an external wastegate. I have seen an old Garrett white paper where they took a T28 and slightly divorced the wastegate discharge by building a simple dam in the housing, built a divorced 02 sensor housing and re-introduced the wastegate flow about a foot further downstream and picked up around 18 hp and better response on a 2 liter engine running something like 12 lbs of boost.
Wednesday, February 22, 2017 11:20 AM
Good article
Dan DeRosia
Dan DeRosialink
Wednesday, February 22, 2017 11:39 AM
Mike, that white paper you mention reminds me of the oft-touted aftermarket "divided/divorced downpipe" deals on the WRX/STI.

I also can't help but think that even a good IWG setup like an EFR is probably going to hit choke flow on the wastegate before an EWG setup might, but obviously that depends on a bunch of other factors.
Mike Kojima
Mike Kojimalink
Wednesday, February 22, 2017 12:12 PM
I think a lot of the divorced internal products came out of my internet posting of experiments and SCC and Turbo articles back in the day. That was an exciting time!
Mike Kojima
Mike Kojimalink
Wednesday, February 22, 2017 1:02 PM
The other thing I have noticed in my experimenting over the years is the bigger the A/R, the less difference a wastegate and how you handle the discharge makes.
Wednesday, February 22, 2017 2:23 PM
@mike, yup, larger A/R turbine housing results in less wastegating. So less advantage to external dump.
Wednesday, February 22, 2017 2:35 PM
The difference in power will really come down to two things: how much is wastegated and how free flowing the exhaust system is. If not much is wastegated, not much to be gained. If the exhaust system is massive and does not have cats, say 1psi of pressure drop at max power, then cutting the flow down by 1/3 (1/3 of total exhaust flow is wastegated) would only result in a pressure drop reduction of about half down to 0.5psi. Assuming the turbine is operating at a pressure ratio of 2.5, that's only a back pressure reduction at the exhaust ports of 1.25psi. However, if the exhaust is smaller and has cats, maybe it has a pressure drop of 5psi. By external dump of 1/3 of the exhaust flow, it would reduce the back pressure at the exhaust ports by 6.25psi which is a ton.
Wednesday, February 22, 2017 5:15 PM
So mike, routing an external wastegate back into the downpipe is better than dumping it to the ground?
Wednesday, February 22, 2017 6:12 PM
@soybeanboy, re-routing the external wastegate back into the downpipe will be quieter and less smelly, but not make as much power as dumping.
Thursday, February 23, 2017 8:29 AM
The mighty SR20 in project G20 made ~225whp at stock boost with stock cams and a divorced internal wastegate on the puny T25. Most guys running the stock O2 housing were in the 200whp range, so there are obviously benefits from divorcing an internal wastegate.
Mike Kojima
Mike Kojimalink
Thursday, February 23, 2017 8:35 AM
Especially when you are wastegating so much with a little turbo.
Thursday, February 23, 2017 2:21 PM
Always wondered about that: do you wastegate more or less with a large/small turbo? Obviously, the less boost you're trying to make, the more the gate is open...
Thursday, February 23, 2017 5:07 PM
@Mike - To clarify your statements, you're saying you increased horsepower over an external wastegate setup by dumping the wastegate pressure downstream of the turbo?

I'm asking because that sounds a lot like a 'Coanda Effect' happening.

Your Image
Thursday, February 23, 2017 7:06 PM
@Rockwood, the larger the turbine housing, the less you wastegate because the larger turbine housing allows for more mass flow through the turbine wheel for the same shaft speed of the turbo.

@ginsu, Mike is talking about your typical Garrett T25/T28 5-bolt turbine housings. Instead of having the wastegate flow merge back with the flow coming out of the turbine wheel in basically an uncontrolled manner dumping into a big volume, the wastegate flow is directed into a tube that merges with the turbine wheel flow further downstream in a controlled and narrow merge angle.
Friday, February 24, 2017 9:54 AM
What is your opinion on a twinscroll external setup where the two scrolls are joined to one large gate, with a divider right up to the gate to theoretically separate pulses? Or a twin scroll external with a wastegate welded right to the turbine housing? My impression is that it would be hard to keep the pulses separate thus negating the benefit of the twin scroll, but to what extent im not sure.
Friday, February 24, 2017 10:45 AM
@Van, in theory, using a single external gate on a twin scroll setups where there is a divider wall up to the gate works. That's how all the twin-scroll internally wastegated turbine housings work after all. A single wastegate flapper covers the internal wastegate hole which has a divider in the middle.

Now, there is a sensitivity to the gap between the flapper surface and the divider wall. If there is a gap between the divider wall and the flapper, then it allows cross-talk between the two scrolls. I can't tell you what the sensitivity is though, that's confidential :)

Personally, I'd just do twin external gates to ensure performance. Or if you have a damn good fabricator who can hold very tight tolerances, then yeah, feasible to go with the single gate on a divided merge tube.
Post Comment Login or register to post a comment.

MotoIQ Proudly Presents Our Partners:

© 2018 MotoIQ.com