Next presentation is Kinder Morgan. Kim's going to do a presentation for us. Please note we can do them both vocal and let's also try to use Slido. If you'd like to text in your questions, there are instructions on the table that you can call up Slido, punch in the #MEIC2019, and then Kinder Morgan, and text in your question. This is a family event, so if you send anything naughty, it will be deleted. Pleased to have Kinder Morgan, a firm known well. They rank number 3 on our new Midstream 50 that will be coming out in the next issue of Midstream Business Magazine. Kim?
Thank you very much. Good to be here after five years. Everybody. We're going to start just at a very high level in the presentation, looking at global energy demand. If you look at the 2018 IEA World Energy Outlook, it shows that natural gas and petroleum demand growth for years to come. You can see on the right-hand side that that's driven by growth in the developing markets. You look at the demand from India, and it's expected to double over the period by 2040. If you look at China, it's expected to become the largest importer of natural gas and oil. Yet still, by the time you get to 2030, you have 650 million people that still lack electricity. The population growth and the urbanization and the economic development in these developing companies create tremendous demand for energy.
Where is all the supply going to come to meet this demand? A large part of it's going to come from the U.S. Looking at the U.S., the U.S. is expected to account for over 50% of the expected growth in supply between 2017 and 2025. We've had a tremendous revival of the energy business here in the United States. You look at our approved reserves, they've doubled over the past 10 years. By 2025, the U.S. is expected to supply one-fourth of every MCF of gas and one-fifth of every barrel of oil. What you have is tremendous growth in supply in the U.S., and that's moving to feed the developing countries' demand. There's a real opportunity in the U.S. to build infrastructure to get the volumes to the coast. Kinder Morgan is very well-positioned to do that.
We've got an unparalleled and irreplaceable footprint that we've built over the last 20 years. We have the largest natural gas transmission network. We're the largest independent transporter of refined products. We're the largest independent terminal operator, and we're the largest transporter of CO2. Our strategy, we focus on owning stable, fee-based assets that are core to the energy infrastructure. Then we try to maintain financial flexibility, and that involves maintaining an investment-grade credit rating and ample liquidity. We're very disciplined in the way that we allocate capital. We have a high return hurdle, about 15% on average unlevered after tax. The assumptions that we assume in the underlying cash flows and the terminal values are also conservative. The goal at the end of the day is to enhance shareholder value over the long term. That involves pursuing attractive projects.
That involves growing our dividend, that involves share repurchases. It also happens because we have a very highly aligned management team. We believe KMI is a core energy infrastructure holding. We've got a greater than $40 billion market capitalization. We're one of the 10 largest energy companies in the S&P 500. We've got investment grade rated debt. We've got a BBB flat rating. We've got a 5% current dividend yield, and I'll show you a little bit later, that's more than twice the dividend yield of the S&P 500. We're growing that dividend by 25% this year and expect to grow it by another 25% in 2020. We've got a $2 billion share repurchase program, of which to date, we've spent about $525 million. We generate enormous cash flow.
If you look at the last three years, 2016, 2017, and 2018, we've generated almost $10 billion in cash flow in excess of our dividends. If you want to look at it on a GAAP basis, if you look at CFFO, which is not a perfectly comparable measure, but some people prefer to look at it that way, we've generated $10.5 billion of cash flow from operations. In 2019, as you can see on the page, we'll generate $2.7 billion of cash flow in excess of our dividend. That just gives us tremendous financial flexibility to generate value for our shareholders. Now, that cash flow that we're generating is very stable. If you look at it, approximately 96% of our segment cash flow is take-or-pay, fee-based or hedged. Look at how the cash flow breaks down, 66. Two-thirds of our cash flow is take-or-pay.
That means that people, our customers pay us whether or not they use the capacity. It's like renting space in a building, right? You pay the rent whether you're occupying the space or not. 25% is what we call other fee-based. What that means is, the price doesn't change. We don't have commodity exposure. What we're really at risk is we have volume risk, not price risk, only volume risk. If you look at the underlying nature of those businesses, 10 of the 25% is coming from petroleum products or products pipelines. Okay. Petroleum product demand in the U.S. is very stable. It typically grows at 1% to 2%. If you have an inflation, maybe it's flat or slightly negative, but very stable. It typically tracks demographic growth. Natural gas pipelines is another 10% of that 25%.
Again, here you don't have price risk. It's volume risk. Here it's primarily on our gathering and processing assets. The key here is making sure that you are in very economic basins. The basins where we have exposure are the Haynesville, the Bakken, and the Eagle Ford. Very economically attractive basins to produce. About 5% of the 25% is primarily in our terminals business. Here we have very highly utilized liquids terminals where we're storing gasoline, diesel, jet fuel. Typically, we have 95% plus utilization in these terminals. Well, these are the ancillary fees that our customer pays us. They pay us a monthly fee to be in a tank. When they do certain product moves, et cetera, they're ancillary fees that they pay. These are ancillary fees that we earn at very high utilization liquids assets.
It's a requirements contract on our petcoke and our steel business. Petcoke is what the refineries produce when they run. Refinery utilization in the U.S. has been very high and expected to be very high for the future. Very stable cash flow coming from the 25%. That's other fee-based. We've got 5% that's hedged, and that's in our CO2 business where we're producing some oil. That cash flow is largely hedged. Here you have both price and volume risk. On the price side, we're hedged on the near term. You don't really have much near term exposure. On the volume risk, these are reservoirs that we have been in for years. This is tertiary recovery. You've already gone through the primary, the secondary recovery. You've got a lot of knowledge about these fields.
We've been within 1% of our budget over the last 9 or 10 years. Able to call our shots on that business. You've got 4% that is commodity based and unhedged. People ask us, "Well, you generate enormous amount of cash flow. How do you think about allocating that cash flow?" Here are the different options that we have. We can allocate it to pay down debt. We can allocate it to increase the dividend. We can allocate it to capital projects or to share repurchase. When you're looking at it for 2019, we have achieved our long-term target of four and a half times debt to EBITDA. In 2019, we're not allocating incremental dollars to debt paydown. On the dividend, we've communicated what our dividend is for 2019, what our dividend is expected to be for 2020.
We know what the dollars are that are going to the dividend. Really when you start looking at 2019, that leaves capital projects and share repurchase. For 2019, really based on the projects we have, all the capital is going to capital projects. That's because we believe that the projects generate a higher return to our investors than share repurchase. That's because we've got a very high return threshold set to pursue projects. Those are going to generate nice returns for our investors. We think where we find those projects, that the capital projects they take priority over the share repurchase. To the extent that we have excess dollars, then those would be allocated to share repurchase. Allocating between capital projects, share repurchase, that's something that we evaluate, we continually evaluate as the situation and the circumstances change.
It's not something that's one and done. We talked about global energy demand. Take a moment to drill down into U.S. natural gas. The reason we focus specifically on natural gas is about 60% of our business is in natural gas segment. We move about 40% of all the natural gas consumed in the U.S. Natural gas demand, as you can see on the right, is expected to increase from 90 to 119 BCF a day between now and 2030. That's 29 BCF of growth at 32%. Very nice growth in demand for natural gas. You can see the biggest growth is coming out of LNG exports. About 14 BCF a day of incremental exports from LNG facilities over that period of time.
Nice growth coming out of the power sector as coal plants convert to natural gas, out of the industrial sector as more development on the U.S. Gulf Coast uses natural gas as feedstock. That's huge demand for natural gas. Production's coming out of four basins, really. You look at that, it's coming out of the Eagle Ford, it's coming out of the Haynesville, it's coming out of the Permian, and it's coming out of the Marcellus Utica. If you look at all the other U.S., it's expected to decline. Really four key basins, of which we touch all of them, but the demand growth is primarily concentrated on the Gulf Coast, in Texas and Louisiana specifically. We've got significant assets in place there in order to capitalize on that. Right now we've got a backlog of about $6.1 billion of projects, okay?
If you look at that, $4.3 billion is in the natural gas segment, that's about 70% of our backlog. On average, we're investing at about a 5.5x EBITDA multiple. Very attractive returns on the projects that we're pursuing. People always say, "Okay, well, what's beyond the $6 billion? What can you do for me beyond that?" We expect that based on the underlying fundamentals that we see in our businesses, that we can invest between $2 billion and $3 billion per year in new opportunities. Another thing that people frequently ask is, "Well, you've invested a lot of capital over time. How have you done on that capital?" I think this slide will show you that we have done very well. We've achieved attractive build multiples.
If you look at the total capital that we've invested between 2015, these are projects that we completed between 2015 and 2018, you look at the capital invested divided by the year two project EBITDA, we look at year two, primarily because sometimes you have some ramps in the cash flow between the first year and the second year, you get a full year in the second year. The original economics when the board approved these projects, the build multiple was 6.1x. What we actually achieved was 5.9x. We have done better than we expected in terms of the economics that we've achieved on these projects. If you look specifically at natural gas, I think that's relevant for two reasons. One, it's 70% of the backlog, it's a big piece of what we've done historically.
The 70% of the backlog, if you look, I'll show you in a minute, our bigger projects that we have coming up are in the natural gas segment. Here we've done even better. We originally anticipated when the board approved these projects, that we would achieve a 5.8x multiple, what we actually achieved was 5.2x. I think that's very important as you get to the next slide, which is two of our large natural gas projects, which are the Gulf Coast Express and Permian Highway. Both of those projects take gas from the Permian Basin to the Gulf Coast. Gulf Coast Express goes more south, is meant to meet more of some of the Mexican demand. PHP is going a little bit further north and hitting our systems on the Gulf Coast in Katy.
Those two systems, when they're in service, will move 4 BCF a day of gas. Our existing system on the U.S. Gulf Coast on average moves 5 BCF a day, and can peak out around 7 BCF a day. I think, when we're thinking about future opportunities beyond the backlog, there's going to be opportunities to continue to debottleneck those Gulf Coast pipes and move those volumes further downstream to the ultimate demand points. Gulf Coast Express expected to be in service in October of this year. Permian Highway expected to be in service in October of 2020. Turning to the liquid side of the business. While the growth here isn't as significant as what we anticipate on the natural gas side, you've got similar dynamics.
That's you've got demand growth in China and India really driving export demand, and you have U.S. is going to be a significant player in meeting that demand. If you look at our Gulf Coast position, and I think one of the reasons, just so everybody's familiar with this, that the U.S. is going to be a big exporter of petroleum products, is because we've got the most efficient refining capacity in the world on the U.S. Gulf Coast. It's much more efficient than the refining capacity in Mexico and Latin America and Europe. The refiners have continued small incremental expansions. Some of them have done larger ones in order to be able to export more product. Our position on the Gulf Coast, we've got 43 million barrels of total tankage capacity. We're handling about 15% of the exports today.
We've got 20 inbound pipes. A lot of those inbound pipes are coming from the refining capacity. We've got 15 outbound pipes. We've got 12 barge docks, 11 ship docks. That puts us in a very nice position to be able to export the incremental refinery production. When you think about beyond the backlog, there's going to be incremental opportunity potentially coming out of the Permian, for potential pipe three, as people say. There is potential for incremental volumes coming out of the Haynesville. There's opportunity to invest in the Bakken as that basin continues to grow. You saw the 14 BCF a day on the slide of incremental demand from LNG facilities. I've heard estimates, and I find this crazy, and I'm not sure I believe them, but estimates as high as 35 BCF a day of LNG exports off the U.S. Gulf Coast.
The number I showed you earlier was the 14 BCF a day increase, but that would just take it to 17 BCF. There are numbers out there that could double that. I think great opportunity to be in the natural gas space right now. If you look at ICF, what they're estimating, $800 billion of North American energy infrastructure investment that is required between now and 2035. Very nice fundamental backdrop. We call this slide the Tale of Two Cities. If you look on the one hand, drill down into the S&P 500 companies, when you look at companies that have debt to EBITDA less than five times, that are investment grade, that are of a decent size, so market cap greater than $35 billion, that have nice projected earnings and dividend growth and have a dividend yield, so attractive dividend greater than 4%.
You get to one company. Huh. Amazing how it works out like that sometimes. If you look at KMI's valuation relative to the average S&P, we trade at a discount to the 12.4x multiple of the S&P. Even more dramatic, I think, is when you look at it on the dividend yield, where our dividend yield is more than double the S&P 500 today, and we've got another 25% increase expected that we're projecting in our dividend for 2020 over 2019. Just to summarize, I think KMI is a compelling investment opportunity. We've got very strategically positioned assets. They're generating $8 billion of 2019 adjusted EBITDA. That cash flow is 90% take-or-pay or fee based, and we've got a 25% increase in our dividend this year. We got another 25% expected in 2020.
We've got a lot of flexibility because we're funding our CapEx with existing cash flow. We've got a management team that's highly aligned that altogether owns about a 14% stake in KMI. We've got an active buyback program. With that, I will be happy to take questions. There is some questions on the screen. All right. The first question is about Can you all read that, or I can read it to you? Yeah, I'm asking, can you see that from the back of the room? Okay. Everybody's got 20/20 vision. That's great. Now that the strategic review of KML ended in no sale, how viable is KML as a standalone company? I'd say KML is a very viable standalone company. It's got about $200 million in EBITDA. I think we have some good growth projects that we're looking at there.
It's got a great Edmonton liquids terminal position that we have built from scratch over the last 10 years. There's potential to expand that. We've got a potential for about a 1.8 million barrel expansion at our BTT terminal. There's potential for some smaller expansions, like for some blending, for pipeline interconnects. They continue to bring us opportunities on Vancouver Wharves, which is a bulk terminal on the coast of British Columbia. Those tend to be chunky projects, to $100 million, $200 million, $300 million projects. We haven't seen any so far that have come to fruition, but I think, one of these days, one of those is going to hit, and it's going to be underpinned by long-term contracts. I think we have some headwinds coming on KML. We've got some contractual rollovers. I think that is known by the market.
I think we've got some opportunity to backfill there, and I think we've got some opportunities for growth that we don't currently have under contract, but I think we have a fighting chance to get some of those deals done. Second question. How do you feel about a potential sale of the CO2 business? For those of you who aren't familiar with our CO2 business, there we do tertiary. It's about 7% of our overall business, so not a huge % of our overall business. There we do tertiary oil recovery. We inject CO2 into the reservoir. CO2 is miscible with oil at certain pressures, and we produce oil. That has been a very good business for us. We've earned very nice returns on capital, and we've even earned reasonable returns on capital when crude prices have been $40-$50.
We've got a management team that we've done benchmarking on, that really knows how to get the oil out of the ground, and who is able to identify unique techniques to do so. We've been able to call our shots on volumes, as I said earlier. In the near term, we largely hedge that. It's a business that is not a huge portion of our portfolio, but generates nice return, that we think we've got very good management team, and that we earn nice returns. That being said, every business that we own is for sale every day. At the right price we will sell any asset that we own. The way we think about that is a couple of ways.
One, if you are going to sell an asset and it's not going to be accretive to DCF per share, then what you have to have confidence in is that you're going to get an expansion in your multiple, that is sufficient to more than offset the dilution that you're taking from the sale. To the extent that we have confidence that that would be the case on a sale of an asset, that's a sale that we would consider. The other thing that we look at is we look at what's the unlevered return to the buyer based on the cash flows that we think the asset is going to generate.
If we look at the set of cash flows and it says that buyer is going to make a 20% return on the cash flows that we know and expect our management team can generate, then that's not an asset we're going to sell. If we see it and we say, "Oh, we think it's going to generate a 2% return." Boom, that's an asset that we'll sell. I think it's an economic evaluation for us when we look at selling assets. PG&E bankruptcy material impact on the Ruby Pipeline. We've talked about the impact of the PG&E bankruptcy. Look, I think PG&E has seen the value of that capacity recently as some of the supply that they get in Northern California has been impacted by some issues on pipelines upstream, and Ruby's been able to fill that void.
It gives them diversities of supply, which gives them security of supply. On average, they're utilizing anywhere 50, 60, 70% of that capacity. That's capacity that when they took out on the pipeline, the commission signed off on. They are expected to hold capacity. We haven't heard anything that would make us think that they're going to reject that contract at this point. At this point in time, everything looks fine. It is a constantly changing situation. I think PG&E pays about $90 million in terms of tariffs to Ruby. It would be a significant impact on Ruby, but as I said, we're cautiously optimistic about them maintaining that capacity. Next question. Internal, our leverage goal. We've hit the four and a half times. Do we feel any pressure from the market to push leverage below four times?
We are comfortable at the 4.5 times. At 4.5 times, we have a BBB rating. We think that gives us sufficient flexibility. That BBB rating reflects the scale, the diversity, the contractual underpinnings of the contracts that we have on our assets. That's a place. The other thing I'd say is, if you look at it from an economic perspective, to go from 4.5 times to 4 times, that's about $4 billion of debt paydown, roughly, that it would take. If you look at cost of capital, you look at where we could issue debt at BBB+ versus BBB. There's not going to be a big difference in our cost of capital on the debt side from taking $4 billion to pay down debt.
If you look at on the equity side, you say, okay, you think you'll trade at a better multiple because you have less leverage. What I would say is, I can't be sure of that. We went from 5.5 times to 4.5 times. If you think there's going to be a benefit in our equity multiple, you darn sure should've seen it then. Since we didn't, I don't have a lot of confidence that going from 4.5 times to 4 times, we would see a big benefit in our multiple, and therefore I don't see that you would get a big benefit in your equity cost of capital. I think, for the time being, and for now, that's the target and that's where we're comfortable.
What is the potential to bring Gulf Coast Express online slightly ahead of time? Gulf Coast Express is expected to be in service in October of this year. That's consistent with what we've been saying since we started the pipe construction. It's in service for our customers when they can utilize 2 BCF a day. Will commissioning take a week or two? Yes, commissioning takes a week or two or three on a gas pipeline. Our customers need this, and we will get it in service for them as quickly as we can, and we currently anticipate that'll be October. Do we benefit when the Waha basis widens? Generally, we do not benefit when the Waha basis widens because we contract the capacity on our pipelines on a long-term basis. That's the model that we pursue, and therefore, we don't have a trading operation or an optimization group.
We generally aren't going to benefit or if we do, it's not really around trading, it's people utilizing our capacity. We don't have a material benefit from that. Would you consider reversing the Cochin Pipeline? Cochin Pipeline has contracts on it until 2024. I think that's something that we can consider at that point in time. What's your outlook from a regulatory perspective, potential expansions on TGP in the Northeast? I think from a regulatory perspective, people are talking about getting projects approved. I think, Ken, are you submitting that question again? All right. I talked about that in the panel discussion. Look, I think it's a tough situation in the Northeast. They need more natural gas, okay? They're burning expensive LNG when it gets cold. That is going to stymie economic growth.
That is a tax on the less fortunate when you could get cheaper sources of energy into your region. From an environmental perspective, they're burning fuel. Natural gas is way more environmentally friendly than fuel oil. The arguments make sense that you should build more natural gas capacity into the Northeast. Common sense is not prevailing in that case. That's all I can say. In the proposed JV with Tallgrass, would we consider taking? On the proposed JV with Tallgrass, that's very early in those commercial discussions. Yes, what that would involve would be converting a portion of WIC and all of Cheyenne Plains from gas into crude service. Where would the gas flow? WIC has a parallel line in the same ditch, we'll just move the volume that flows onto one line and take one out of service.
On Cheyenne Plains, there are a couple of customers there, we think between Kinder Morgan and Tallgrass that we would be able to provide alternative comparable service to those customers. Is PHP delayed at all? Expected in-service date on PHP is October of 2020. The question about delay, I think is getting at we in the Hill Country in Texas. That pipeline runs from Permian Basin. It runs just south of Austin and into Katy. In the Hill Country and around Austin, just west of Austin, we have seen some of the landowners protesting about our ability to exercise eminent domain. When we originally contemplated this pipeline, we think we picked the best route, the route that is most direct, the route that impacts the least number of landowners, and the route that has the least environmental impact.
We think we've got the right route. We also knew that going this route was going to be tougher than what we did on GCX, where we went further south because the endpoint on GCX is Agua Dulce, which is further south. We knew that it would be a little more challenging, we budgeted more time and more dollars on the right of way to be able to get PHP done. There's been a lawsuit filed. There's also legislation. The legislation really is about future projects. When you think about the outlook for the future, I think that what gets done in the legislature, our current expectation is that it'll be something that is reasonable and that the industry can live with. On the lawsuit, they've filed for an injunction. Don't think that that is likely to be granted.
If it was, I think that's something that we would appeal. I think at the end of the day, that's something that we likely win. Now, when we budgeted more time, we did not budget it to take something to the Supreme Court. That could add a few extra months, but it's not going to be years or something like that. That's the state of that pipe. With that, I'm out of time.