Two weeks ago RBN Energy posted Part 1 of a series looking to answer the question – ‘Are we likely to run into storage issues with NGLs in PADD 1 while we are waiting for infrastructure and demand side projects such as export terminals and petrochemical facilities to be built out?’ The author assessed growing supply and demand mismatches, how production will move between regions, and set the stage for today’s blog where the author examines the need for and availability of NGL storage capacity in PADD 1.
Kelly Van Hull, RBN Energy
First here is a brief recap regarding NGL production growth expected in PADD 1 over the coming years from part 1 of this series, Breakin' It Up & Breakin' It Down – PADD 1 NGLs Storage Capacity: Supply and Demand. A full read of part 1 will likely be very helpful in providing context and information on infrastructure we will be referring to in today’s blog. Historically, NGL production from PADD 1 has been very low, and the need for storage has also been relatively low to keep the market balanced, but as we previously discussed in depth in the Big Surge Comes to Whoville – Northeast NGLs to Increase Six Fold between now and 2015 nearly 4.7 Bcf/d of additional cryogenic natural gas processing capacity, along with 500 Mb/d of fractionation capacity and 500 Mb/d of NGL pipeline takeaway capacity are due to come online in the Northeast to support growing Utica and wet Marcellus production. Once the above mentioned infrastructure comes online, by 2015 NGL production from the Appalachian Basin could exceed of 400 Mb/d. One way or another all of this production is going to have to find a home. Some of it will find a home locally, some will find a home in Canada, some will find a home on the Gulf Coast and some will get exported via the Marcus Hook terminal on the Delaware River.
Before we start breakin’ it down and assessing storage requirements going forward for each individual purity product we are going to take a walk back in time and look at historical PADD 1 NGL stock levels over the past 10 years.
Figure #1, Source: EIA
A lot can be gained from looking at Figure #1 above. The graph on the left shows ending stocks for propane and normal butane, with propane swinging from 1 MMbbls to 6.5 MMbbls and normal butane swinging from 0.5 MMbbls to 3.5 MMbbls. On the other hand, as shown on the graph to the right, storage requirements for isobutane swing less than 200 MBbls and storage for pentanes plus (natural gasoline) is less than 50 Mbbls. There is no ethane storage since there is little or no ethane currently produced in the PADD 1 market.
Getting back to propane and butane, note the degree of seasonality in the storage requirements of the two products. As discussed in part 1 of this series, demand for propane is highly seasonal given the primary demand source is for heating during the colder winter months. The demand for normal butane is equally as seasonal, although the overall quantity demanded is significantly smaller. For normal butane the seasonality of demand is driven by the motor gasoline blending season, also discussed in part 1 of this series. These seasonal demand cycles for propane and butane are unlikely to change any time soon. Also evident in the left chart above is the declining demand, and storage needs for normal butane and the impact of changing weather conditions on propane demand and stocks. As a result of mild weather in 2011 and 2012, propane stocks hit their historical maximum at 6.5 million barrels in Sept 2012. Thanks to some very cold weather moving in over this past winter, propane stocks in PADD 1 had fallen back down to 1.3 million barrels as of the EIA’s March 2013 data release.
Normal butane stocks in PADD 1 have fallen in recent years due to refinery closures and refinery production decreases in the region. Less normal butane is being produced at refineries, and less is needed from gas plants for blendstock, see Figure #2 below for net refinery and blender inputs of normal butane since the beginning of 2009.
Figure #2 Source: EIA
While demand for normal butane in PADD 1 has been on the decline in recent years, this doesn’t preclude the need for additional normal butane storage in the coming years as gas plant supply increases. Normal butane as a percentage of the barrel is far less than either ethane or propane in PADD 1 and as a result the gas plant production increase of normal butane between 2012, 6 Mb/d, and 2015 is only 20 Mb/d. This amount should be manageable via existing storage, rail, proposed pipeline projects, and exports. During 2012, an average of 4 Mb/d of normal butane was exported from PADD 1 to Canada.
The picture gets more complex when evaluating the need for additional propane storage in the region. Recapping from part 1, historically propane demand in PADD 1 far exceeded historical gas plant and refinery supply of propane from the PADD which as of March 2013 were 38 Mb/d, and 36 Mb/d respectively. Thus, propane has historically, and continues to flow into PADD 1 via pipeline and rail. Flows into the PADD increase during the colder winter months when propane demand is higher and fall off during the spring and summer as warmer temperatures return, see chart below. Most propane moving from PADD 3 to PADD 1 is transported on Enterprise’s Dixie Pipeline to meet propane demand in the Southeast while most propane moving from PADD 2 to PADD 1 travels on Enterprise’s TE Products line which runs from PADD 3 through PADD 2 and into PADD 1 to meet Northeast propane demand. Propane has also historically been imported from Canada and via ship into PADD 1.
Figure #3 Source: EIA
As can be seen in Figure #3 above, flows from other PADDs swing significantly from summer to winter and also relative to the severity of the winter. During the mild winter in 2011/2012 less propane needed to move into the region to meet demand. In early 2013 the impact of growing gas plant supply of propane can also be seen in the chart. Gas plant production of propane has grown from 16 Mb/d in January 2011 to 38 Mb/d as of March 2013.
Natural gas plant production of propane is expected to grow significantly over the coming years, with production in excess of 120 Mb/d by 2015. In contrast, refinery production of propane is expected to remain flat. As a result of the growing gas plant supply of propane the need for imports and flows from other PADDs into the region to balance the market will be reduced. Evidence of the falling need for imports can be seen in the chart below.
On the demand side, propane will get an additional demand boost from exports in 2014 via the Sunoco/MarkWest Mariner East project and the Marcus Hook export terminal that has an anticipated initial capacity of 70 Mb/d and will be able to transport propane by the second half of 2014.
Growing supplies of propane will also be able to move out of the region via the Williams/Boardwalk Bluegrass pipeline as part of a y-grade (mixed NGL) stream by late 2015 and may also be able to move on the Enterprise ATEX line by 2015 if sufficient commitments are received during the current open season. For more information on Enterprise’s open season to determine shipper support for transporting propane on the ATEX pipeline to Mont Belvieu see part 1 of this series.
Assuming Mariner East, Bluegrass and ATEX propane all happen, and heating demand for propane remains flat there should be enough propane storage in PADD 1 to keep the market balanced. In the short term before those outlets are completed surplus propane will move out of the region via rail during low demand periods and back into the region via pipeline to meet winter heating demand.
Assessing the need for ethane storage in PADD 1 is even more complex than propane. Historically, no ethane has been produced in the region due to lack of local demand. In the future, ethane will be produced from the region both as a result of new takeaway capacity to meet demand from petrochemical facilities in Canada, Europe and the Gulf Coast, and also as a requirement to produce the other NGLs from the region. For more details on growing ethane production in PADD 1 see part 1 of this series Breakin' It Up & Breakin' It Down – PADD 1 NGLs Storage Capacity: Supply and Demand.
By 2015 over 200 Mb/d of ethane could be produced from the region, but no incremental local demand will have developed by that point so all of the ethane will be leaving the region via pipelines planned and under construction by Enterprise and Sunoco/MarkWest. Sunoco/MarkWest’s Mariner West Project is the first due online, with an expected in service date later this year The Mariner West Project is a 50 Mb/d ethane only pipeline to move ethane from the Houston fractionator in Western Pennsylvania to the Sarnia, Ontario petrochemical market. Also proposed to move purity ethane, and possibly propane (given Enterprise’s recent open season announcement), is Enterprise’s 190 Mb/d ATEX pipeline which will move products from the Appalachian Basin to the Texas Gulf Coast. In addition to the ethane only pipelines just mentioned Sunoco and MarkWest have also proposed the Mariner East project, mentioned above, to move ethane and propane to the eastern seaboard for export. In total these three projects provide for 280 Mb/d of ethane takeaway capacity. Given adequate pipeline takeaway capacity will be in-service by 2015 to move all 200 Mb/d of incremental ethane production there will not be a significant need for ethane storage in PADD 1 to balance the market through 2015.
The ethane storage picture could get more complicated when you get out to 2019 with the possibility of a new world scale cracker being built in the region. Shell has been the most prominent in proposing to build a new cracker in Appalachia. Should Shell’s facility or some other plant be built it could demand 80-85 Mb/d of ethane. That will be a good source of demand for additional ethane that could be produced in the region by 2019. But the catch is that ethylene plants are notorious for downtime, both scheduled and unscheduled. So where does the ethane go when the plant goes down? This could be handled via pipeline space or storage capacity, but the storage would have to be underground in a cavern as storing purity ethane in this kind of quantity above ground is not feasible. Inergy has existing salt storage caverns for storing NGLs in New York and Sunoco Logistics has five underground storage caverns at Marcus Hook, and there is underground NGL storage in Sarnia in Canada at the far end of the Mariner West project. All of this storage is currently used for existing customers. Whether it would make sense to use some of this existing storage to handle cracker disruptions is a big open question for Shell or any other developer of a steam cracker in the region.
Both isobutane and natural gasoline are also produced and consumed in PADD 1, but as noted above, storage requirements for these two products are minimal relative to propane and normal butane and this will remain the same in the future as production growth for these products is much smaller.
In the end there will not be a need for a large amount of additional NGL storage in PADD 1 as excess supply will be leaving the region as quickly as possibly via planned pipeline and export projects. Yet, as with many things in the energy world, systems do not always run fully optimized and things don’t always go as planned so companies like Chesapeake, with partners M3 Midstream and EV Energy Partners are developing NGL storage capacity to go along with their other infrastructure. At the partnerships’ planned large midstream services complex in eastern Ohio there will be 90 Mb/d of fractionation capacity, a rail-loading facility and an initial NGL storage capacity of 870,000 barrels.
While there will likely be hiccups along the way and some deviations from historical NGL stock trends as the proposed infrastructure comes into service in PADD 1, and NGL production ramps up, ultimately the PADD 1 NGLs storage picture will likely look very similar to history and the charts at the beginning of this blog. If a new world scale cracker is built in the region in 2019 or beyond then ethane storage will have to be handled when the cracker goes down. Until then we do not expect additonal capacity will be needed.
That’s it for Breakin' It Up & Breakin' It Down for PADD 1 NGLs Storage Capacity: Storage Requirements. Stay tuned as we take a look at the other PADDs over the coming weeks.
About the author
Prior to joining RBN Energy, Kelly worked for BENTEK Energy where she built supply/demand models covering all US NGLs. She was also the developer, lead analyst, and writer for BENTEK’s North American NGLs Market Call report specializing in supply and demand fundamentals for natural gas liquids markets, and the lead analyst and writer for BENTEK's NGL Supply Report. Prior to joining BENTEK, Kelly was an analyst for a midstream company developing greenfield projects in the Williston Basin. She holds a BA in Economics from the University of Michigan.