Marcellus Opportunity – Why We Should See More Chemical Investment

Print Friendly, PDF & Email


Graham Copley / Nick Lipinski



September 9th, 2014

Marcellus Opportunity – Why We Should See More Chemical Investment

  • As the supply of natural gas rises in the Marcellus, and soon also the Utica, the gas is having to move increasingly longer distances to find customers (note Dominion/Duke pipeline announced last week). This is resulting in decreasing wellhead values for natural gas and NGLs.
  • Production costs in the Marcellus keep dropping and with good co-product values for propane and butane, wet gas is still the preferred route. Natural gas and NGL surpluses in the Marcellus should grow for the foreseeable future and all current estimates put production well above any expected local demand.
  • Depressed local values for natural gas and ethane can be partly offset by pipelines to either more distant customers or export facilities, but the cost of transport will still lead to a significant price discount in the region. Today – should a facility exist – the cash cost of producing ethylene in the Marcellus would be around 5-6 cents per pound lower than in the US Gulf.
  • Separately, we are seeing increasing trouble with transportation in the US; infrastructure cannot keep up with increasing rail network demand, new tank car safety standards, and constraints on the roads. Plastic production closer to a key customer base could give producers in western PA or WV a further significant cost advantage over their US Gulf counterparts.
  • To date there are three announced ethylene facilities for the Marcellus – Shell, Braskem and a small project from Aither Chemicals. We think there should be more and we think that the smarter US players will start to get involved quickly, for both offensive and defensive reasons.
  • Risks all surround the possible value of ethane in Marcellus and its availability. If natural gas prices in Marcellus move to US Gulf less transportation costs (they are well below this today), you get the costs suggested in the last bar in Exhibit 1 – still better that a new build in the US Gulf, by around 10%.

Exhibit 1

Source: Mid-Stream, IHS, SSR Analysis


In the last few months, natural gas production in the Marcellus has surpassed the demand for natural gas in the states surrounding Marcellus. This means that additional supply, of which there is expected to be plenty, will be transported further to find a market – further depressing the value at the wellhead. Over the last few days we have seen an announcement from Duke Energy and Dominion Resources that they will spend $5bn to build a pipeline linking the Marcellus with North Carolina via Virginia – a distance of 550 miles. It is likely that for the foreseeable future, natural gas values in Marcellus will be lower than in the US Gulf, particularly if the US Gulf has an LNG export option and the North East does not. While the industry rushes to build infrastructure to move both the natural gas and the NGLs to consumers, it is likely that we will see periods of logistic constraints through 2015 and 2016, when natural gas in Marcellus could sell for pennies at the wellhead.

The money is being made today in the LPG components of the “wet gas” and this is reducing the Finding and Development (F&D) costs for natural gas in the region such that local E&P companies are making money with local natural gas prices below $2.00 per MMbtu. Ethane in Marcellus is likely to be available at a discount to fuel value equivalent in Marcellus, and so consequently also significantly lower than the US Gulf. While compelling today, for forward planning purposes, this may not be a sustainable advantage, as with sufficient infrastructure it would likely cost no more than $0.50 per MMBTU to move natural gas from the Marcellus to the US Gulf. This is probably a more appropriate conservative approach to analyzing the opportunity. Taking this view, we still see a competitive advantage making ethylene in the Marcellus, but probably not enough to justify the risk on its own.

However, there is a significant movement of polyethylene from the producers in the US Gulf to the mid-West and the North East. We estimate that as much as 30% of US polyethylene demand lies within 500 miles of Natrium, WV – one of the sites being considered for ethylene investment – this would be around 5 million tons. Today, because of a shortage of rail infrastructure, rail cars and truck drivers, shipping costs from the US gulf to the mid-West and the North East are rising. Investments will likely be made to alleviate some of these constraints but infrastructure constraints are likely to be a long-term problem in the US and it is unlikely that shipping costs will fall much below recent averages. Local producers of polymer should have a significant logistic advantage over their Gulf Cost competitors.

So the question of the day is; with lower production costs and lower logistic costs –“why isn’t everyone looking to build ethylene/polyethylene in West Virginia or Eastern PA?”

Today we have three companies looking at projects in the region, but no-one has permits yet and final capital decisions have yet to be made. However, none of these companies has a “strategic” interest in producing polyethylene in the US Northeast. At the same time, none of the companies currently shipping polymer from the Gulf appear to be looking, and they are the ones who could get squeezed out if there is significant investment in Marcellus; forced to increase (lower return) exports to make up for lower domestic demand. This is a longer-term issue and will not upset market dynamics for the next 4 years, but most exposed would be Dow Chemical, Exxon Chemicals, CP Chemicals and LyondellBasell.

Gas production in Marcellus

The Marcellus formation has seen average annualized production growth of 22.5% from 2007 through 2013 – Exhibit 2. Production for 2014 has already outstripped all of 2013’s production, indicating continued growth at this elevated rate. Production growth in excess of 20% per year is anticipated for the next several years. As shown in Exhibit 2, the increased production is being achieved with a lower rig count, indicating a rising production per rig and suggesting lower F&D costs. Range Resources and Cabot Oil & Gas have published their F&D costs for the last several years and the trend – as shown in Exhibit 3 – is clear.

Exhibit 2

Source: Energy Information Administration

Exhibit 3

Source: Range Resources, Cabot Oil & Gas, SSR Analysis

Projections for Marcellus estimate an additional 3 bcf/d in 2014 plus 2 bcf/d in 2015. Production will likely top 16 bcf/d before the end of this year. It is because of this growth in production and the need to move the gas further and further to find a market (coupled with pipeline restrictions) that prices in the Marcellus have collapsed and diverged from their historically tight relationship with Henry Hub prices. As illustrated in Exhibits 4 and 5, it now becomes a question of logistics and distance. Marcellus is producing more gas than is needed in the surrounding states and E&P companies are only just getting to grips with the Utica shale, which sits below and to the Northwest of Marcellus and looks to be bigger! Industry is convinced that the surpluses will grow and will be sustained, as illustrated last week by the aforementioned announcement from Duke Energy and Dominion Resources that they have plans for a new $4.5-$5.0 billion pipeline. This will allow Duke to access Marcellus gas, instead of or in addition to Gulf Coast gas as it continues to shift its power generation mix towards natural gas.

Exhibit 4

Source: Energy Information Administration

The diverging prices of Marcellus gas and Gulf Coast gas are shown below. Current pricing in the Marcellus is below the cost averages suggested by Range but above the Marcellus specific cost suggested by Cabot (Exhibit 3). This lower cost likely reflects the higher values of propane and butane and the benefit of wet gas versus dry. While ethane is clearly in surplus and selling at a discount to its fuel value in the US Gulf and at Conway, propane, butane and pentane remain supported by higher crude oil prices. Fractionator overall margins remain high – disposing of the ethane is the current problem and it is expected to remain a problem for some time.

Exhibit 5

Source: EIA, Bloomberg, SSR Analysis

Equally important, Marcellus, and going forward Utica, are increasing natural gas availability in a region of slow natural gas consumption growth. Power generation conversion to natural gas continues but not at a rate fast enough to keep pace with supply and the region will increasingly rely on pipeline infrastructure to move gas further and further to each marginal consumer. By contrast, the growth in production at both Eagle Ford and Haynesville (in the South) is increasing supply in a region of relatively fast demand growth, in part driven by the demographic and business shifts in the US and in part because of the expected start-up of at least one LNG export terminal.

Consequently, we can see reasons why natural gas prices in Marcellus should remain depressed and reasons why they should remain supported in the US Gulf.

Ethane Looks The Same

Both the Northeast and the Gulf Coast are seeing investment in ethane export terminals, and while this may support pricing, we believe that it is more likely to do so in the South where there are multiple users and because there could be multiple buyers for the export market. Without additional export capacity in the Northeast it is unlikely that there will be much competition for export. We think that only Ineos has it right today – the sustained ethane surplus is likely going to be in the Northeast and not in the Gulf; Ineos should have a pricing and transportation advantage versus others looking to move ethane to Europe or India.

In Marcellus it looks likely that marginal ethane will be priced at a fuel value discount to a depressed local natural gas price, or alternatively at some sort of netback based on US Gulf pricing less the cost of getting the ethane or the NGL stream to the US Gulf.

Based on current pricing an ethane buyer in Marcellus could pay half the price for ethane compared to a buyer in the US Gulf. The only reason for this to change would be if local ethane demand in Marcellus grew to absorb local supply. Given the NGL volumes available in Western Marcellus and likely to come out of the Utica field, we think this is highly unlikely. Based on projections by Enterprise Products, the Marcellus region could have enough ethane by 2020 to feed 5-7 world scale ethylene units.

Today, at depressed US Gulf ethane prices, cash costs of ethylene in the Marcellus (should a world scale plant exist) would be as much as 5 cents per pound lower than costs in the US Gulf. We struggle to see how this cost advantage could get much worse for the foreseeable future and can see scenarios where it would get better.

Rail/Truck Cost Inflation

Costs of transportation have been rising recently, a trend indicative of improvement in the domestic macro-economy but also influenced by infrastructure constraints. The average railroad company made revenue of 4.83 cents per ton mile in 2Q2014. This represents the highest revenue ever generated per ton mile and coincides with the second highest number of carloads transported (12,002,330 in 2Q2014) on record as well (Exhibits 6 and 7). Trends driving these records are primarily capacity, congestion and capex. This is to say that rail transport has never been in higher demand with capacity virtually maxed out and capex limited in its ability to immediately improve freight handling. The increasing number of carloads is illustrated below. Last quarter it recaptured the 12,000,000 cars per quarter level only seen once in 2006.

Exhibit 6

Source: Bloomberg, SSR Analysis

Exhibit 7

Source: Bloomberg, SSR Analysis

As a result of record shipments, congestion is lowering efficiency with average rail velocities down (Exhibit 8) and dwell times up (Exhibit 9).

Exhibit 8

Source: Bloomberg, SSR Analysis

Exhibit 9

Source: Bloomberg, SSR Analysis

Further increasing costs are proposed Department of Transportation regulations requiring cars carrying flammable commodities to be retrofitted or replaced that will be enforced in 2016. The impacts of this potential legislation are already being seen, for example in an announcement by Dow Chemical that it plans on selling its railcar fleet rather than upgrading it. This has the added impact of taking some of the older rail cars out of service almost immediately.

The major bottlenecks for rail appear to be in the Chicago and south of Chicago area, where increasing rail transport of crude oil from the Bakken is overloading the system. Further complicating logistical matters is a backlog of agricultural products both delayed from earlier this year and expected from record crops after a very favorable growing season.

Trucking – Also Causing Headaches

Truck transport is also a growing liability for manufacturers, with revenues per mile of trucking companies at or near all-time highs. Trucks are transporting goods in record volumes with total tonnage having eclipsed previous all-time highs in 2013 and little sign of slowing in the first half of 2014. The ATA creates an index indicative of the number of tons of freight moved per month, effectively measuring how much freight trucks are moving. That index is shown in Exhibit 10 below.

Exhibit 10

Source: American Trucking Association, SSR Analysis

Demand for trucking is also illustrated in orders for new trucks and the current backlog.

The backlog stands at its highest level ever and orders for new trucks are close to an all-time high, having come down only modestly from their peak in January (Exhibit 11). While the subject is well outside the scope of this report, we are not discussing the added complication of overall weak infrastructure in the US – roads, bridges and rail. The sector has seen below necessary investment for decades and there appears to be little appetite either at the federal level or at the state level in many states to make things better. Increased truck volume will only make thing worse.

Exhibit 11

Source: Bloomberg, SSR Analysis

Furthermore we have a well-publicized shortage of truck drivers in the US – this can be fixed but probably not without cost inflation. When it is considered that there are an estimated 3 million+ heavy duty trucks in the United States, the dearth of drivers, illustrated below (Exhibit 12), becomes even more striking. It also follows that given high demand for drivers, wages that have been largely falling or stagnant over the past several years should rise thus increasing transportation costs further.

Exhibit 12

Source: Bureau of Labor Statistics, SSR Analysis

Exhibit 13

Source: American Chemistry Council Economics & Statistics Department

With roughly 90% of plastic resins moving across the country via these two modes, a reduction in the distance the final product must travel could mean significant cost savings.

Polyethylene demand in the Mid-West and North East

Exhibit 14 shows a breakout of plastic resin production by state with data from 2012. The top five states include two from the Marcellus region, namely West Virginia and Ohio. These top five states produce almost two thirds (64%) of total domestic resin. However, this grossly understates the demand for polymers in the region around the Marcellus as significant volumes are shipped from the US Gulf.


Exhibit 14

Source: American Chemistry Council Economics & Statistics Department

Identifying the level of consumption of polyethylene in the region is very difficult, as there is little data in the public arena. However, if we look at a 500 mile radius around Natrium WV, the area includes Chicago, Indianapolis, all of Ohio, all of PA and New Jersey and much of New York, including New York City. We conservatively estimate that 30% of US polyethylene demand is in this region – close to 5 million tons of polyethylene.

By contrast, New York is around 1500 miles from the US Gulf and Chicago is over 1000 miles from the US Gulf.

We focus on polyethylene because it becomes more complicated to make other ethylene derivatives because of the need for other materials. To make PVC, for example, you need a source of chlorine and demand for caustic soda. (Axiall has capacity for chlorine in Natrium, WV). The other possible derivative of ethylene would be ethylene glycol, which only needs oxygen and water as additional raw materials. Given that ethylene glycol has lost market share to propylene glycol as an aircraft de-icing material because of environmental concerns, it is unclear to us how much demand there is for ethylene glycol in the Northeast and Mid-West

Economics Summary

In Exhibit 15 we show a summary of illustrative comparative economics. In our attempt to model production economics in the Marcellus we value ethane in three ways; first based on the same discount over extraction economics that we see in the US Gulf today, second based on a netback versus US Gulf ethane prices less the cost of piping ethane from Marcellus to the Gulf, and third by resetting the natural gas price in Marcellus based on a reasonable cost of transporting the natural gas to the US Gulf. There are plenty of other scenarios that we can envisage, but the three below probably cover the range of possible outcomes.

Exhibit 15

Source: Mid-Stream Business, IHS and SSR Analysis

The assumptions are listed below:

  • All prices are current – natural gas pricing is taken from Mid-Stream Business and overstates the Marcellus price as it reflects the Ellisburg NE hub price which requires some transport from the Marcellus.
  • In the first Marcellus estimate, the ethane price in Marcellus is assumed to reflect the same frac-spread discount that we see in the US Gulf today. The discount is a little higher in Conway today reflecting the more limited market. The discount is unlikely to be much higher than shown above as at some point using ethane as a local fuel makes more sense. Movements in propane and butane prices could influence how much of a haircut producers are willing to take on ethane and these could be different region to region.
  • The pipeline estimate is based on public data reportedly provided by Range Resources in discussion of their ATEX pipeline contract. We have seen similar estimates from other sources and the number is consistent with rates posted in ATEX’s FERC documents.
  • It is assumed that it will cost 10% more to build an ethylene plant in Marcellus than the Gulf – the Gulf has all the infrastructure and you could argue for a higher premium, but the Gulf also has all the building and the wage inflation associated with it.
  • We are assuming that the ethylene plants cost $1.00 per pound to build ($1.10 in Marcellus) and that polyethylene costs $0.60 per pound to build in both locations – we are then assuming a simple 10% charge for capital.
  • Polyethylene costs (above the cost of ethylene) are assumed to be less on a new plant, and less again in the Mid-West and Marcellus because of cheaper natural gas, though this is not the case today in the Mid-West.
  • We are assuming a 3 cent per pound freight advantage.

On this basis, Marcellus is not a more competitive supplier to the local region than depreciated US Gulf producers. However, it is a much more competitive supplier than new US Gulf facilities. Given that the expectation is that more and more US ethylene moves into the export market as derivatives, as capacity increases, the logical move is to build polyethylene in PA or WV and ship into the export market from the Gulf. This will require collaboration – something we have seen significant resistance to as the recent plans in the US have taken shape.


There are several downside risks to building in Marcellus but they fall into two main categories; anything that could inflate local Marcellus ethane values relative to US Gulf ethane values, and anything that could constrain ethane production in the region. A secondary risk would be that we have the construction cost estimate wrong for the region, given the need to build infrastructure that is largely already in place in the US Gulf.

  • Marcellus ethane prices would rise relative to the US Gulf if there was more demand for the Marcellus ethane than supply – enough local demand, and/or exports and/or capacity to move to the US Gulf.
    • If local demand rises above supply, prices could rise well above those in the US Gulf, with local consumers (ethylene producers) having no alternative.
    • If the demand is for export it depends on the level of export demand and the value in the importing country: the higher the price of oil relative to US natural gas the more a European importer will be willing to pay.
      • If the US price for ethane is set by export pricing (this is what is happening to propane today), the price in Marcellus will depend on whether the marginal export capacity is on the East Coast or in the US Gulf. Marcellus pricing will be lower if the marginal gallon of Marcellus ethane has to be piped to the US Gulf for export.
    • If the marginal gallon is moving by pipe to consumers in the South, Marcellus ethane will be priced based on a netback versus ethane prices in the south.
  • Oil prices falling, and overproduction of natural gas in Marcellus versus demand outlets would be the most likely drivers of limited ethane avialability.
    • Lower oil prices would lower co-product values for propane and butane and lower incentives to export everything – a glut in the US may result in lower natural gas investment in the Marcellus and restricted supply.
      • If the region builds capacity to consume 150 MBPD of ethane and instead of the 300-350 MBPD that is forecast to be produced, the production drops below 150 MBPD, local ethylene producers would simply have to cut operating rates unless pipelines could be reversed and there was sufficient availability elsewhere.
    • Local overproduction of natural gas is likely in the near-term as pipline construction is unlikely to hit completion dates that match perfectly with production changes. Over the next two to three years we could see short periods of real natural gas surpluses in the region as supply surpasses the capacity to move the gas, with very low marginal pricing as a consequence. Any slowdown in the rate of infrastructure investemnt could lead to prolonged period of supply overhang in the Marcellus and this would likely curtail E&P invetment in the region.
      • The risk for an ethylene producer would be volatility of both supply and pricing.

©2014, SSR LLC, 1055 Washington Blvd, Stamford, CT 06901. All rights reserved. The information contained in this report has been obtained from sources believed to be reliable, and its accuracy and completeness is not guaranteed. No representation or warranty, express or implied, is made as to the fairness, accuracy, completeness or correctness of the information and opinions contained herein.  The views and other information provided are subject to change without notice.  This report is issued without regard to the specific investment objectives, financial situation or particular needs of any specific recipient and is not construed as a solicitation or an offer to buy or sell any securities or related financial instruments. Past performance is not necessarily a guide to future results.

Print Friendly, PDF & Email