Tech Talk - Projections 2

It is the end of another year, or more optimistically the start of a new one. Last year I was tempted to make a couple of predictions for the future. And while I can make the case that they were not too wrong, they did not include the drop in oil prices, which has now taken the price of our local gas to below $1.85 a gallon. China has, in recent months, seemed less belligerent about claiming large sections of the China Seas. Whether this has anything to do with the relative success of rigs that have drilled in those waters is something that still remains an unknown.

But it is the changing price of gasoline, itself reflective of the drop in oil prices that is the big news. WTI closed at $53.56 today, and Brent at $57.50 a barrel. Predictions include some who would suggest that the price will continue to fall, until it reaches $20 a barrel, and there it may stay for some time. Well it certainly grabs a headline, but that is about all the value that particular forecast contains. The futures prices suggest that the price has yet to bottom out, though it may be getting close to that value.

Figure 1. Crude oil futures prices (EIA TWIP)

None of the recent news suggests that there will be a further increase in supply to sustain the current imbalance between available supply and demand. Libya is descending even further into a mess, with the oil facilities at the port of Es Sider now being destroyed. The likelihood of significant increases in production and the return to export levels achieved earlier this summer seems increasingly nonexistent. Neither Russia nor Saudi Arabia are likely to increase production, although the latter are continuing to produce the increased volume that they originally put on the market to replace Libyan losses. And so this leaves Iraq and the United States as the key producers who can significantly change the current supply:demand balance in any significant way.

It is probable that, with the agreement between the Kurds and the Central Government now having generated a second payment of $500 million to the KRG that the agreement may be sustained and grow. At present the Kurds are to supply about 550 kbd, of which 300 kbd will travel through the new pipeline to Turkey and thence onto the world market. The rest will be supplied to Baghdad. Meanwhile production in the south (which gets exported through Basra) has seen some increase.

Whether the Kurdish production can increase to over 1 mbd by the end of next year remains open to some doubt, given the ongoing conflict, and the target 6 mbd by the end of the decade for the entire country will likely require changes that the current conflict, which shows no signs of ending, will inhibit.

One of my responses, when the drop in price first started, was to note that the oil supply system has a certain inertia to it. And here I am not talking about the fluctuations in price that one sees in the stock market, and in the price of the crude, but rather in the time that it takes to stop current drilling, postpone future plans and to reduce the production from existing and new developments.

Thus the drop in investment in new production, whether in Russia, Iraq or the United States takes some time to have an impact. Unfortunately for those expecting the price to continue to fall, in the face of the overabundant supply, the situation has changed since historic times, where well production was relatively stable and the oversupply situation was corrected by shutting in production (mainly by Saudi Arabia). Even then it was the perception of the response that drove price rebounds, rather than the immediate reality of the changes.

The system this time is different. The increase in production in the United States has been sustained, and over the last two years has produced more than 2 mbd more than at the start of that period.

Figure 2. US crude oil production over the past two years. (EIA TWIP)

The rig count in North Dakota has already fallen to 170 rigs compared with 187 at this time last year. Concern about the oil price has led companies to cut their investment plans for next years, in some case by 20% so that the rig count is likely to continue to fall. And with the short life at high production values for most wells that will soon affect production. The North Dakota Oil and Gas Division of DMR shows the consequences of this:

Figure 3. Future production estimates from the ND DMR Oil and Gas Division.

The blue line requires about 225 rigs in continuous action, so that won’t happen. By the same token the black line is with no more drilling, and that won’t happen either. The result will be somewhere in between, probably moving the peak out beyond the current projection, but also lowering it as the existing baseline drops with less wells significantly contributing. (Bear in mind it is taking 11,892 wells to sustain current production levels.) But in the short term the line will likely dip down until the price rebounds.

The question now becomes how soon that drop in US production will become evident, and have some impact. I doubt that it will be before June of 2015.

On which note may I wish all readers a Happy, Healthy, Successful and Prosperous 2015.

Waterjetting 28c - using steel as an abrasive

One of the considerable disadvantages in using garnet and other similar minerals as a cutting agent in abrasive waterjet cutting is that the particles fragment during the mixing process, and when they hit the target. As a result (as noted in the last post) less than 50% and often much less than that can be usefully recycled. The distinction in adding the word usefully relates to the need to remove the finer particles from the recycled stock since that does not cut very well.

But what if we used an abrasive that is not degraded in the mixing process, and further one that can be more easily separated from the cuttings and spent water? The candidate is steel, which can be formed into small particles that do not degrade in size as they move through the mixing chamber, and generally hold shape even after they have hit the target. Steel also has the advantage that it can be magnetically removed from the jet stream as the flow is collected, and with no significant degradation in size it can then be readily recycled. In cases where we have monitored the recyclability of steel shot, we were able to re-use it more than fourteen times without seeing any degradation in performance. Re-using it this many times more than offsets the increased price of the original material, and will, in a short time, also pay for the relatively low costs of a magnetic separator.

Unfortunately it is not quite that simple a choice. There are a number of other considerations, which must be addressed to make the system work effectively, some of which may make the process too expensive. Three of the areas that need to be addressed form the subject of this post.

The first comes about as a result of the shape of the particles, and their retained mass and velocity on leaving the focusing tube. More than most other abrasives steel retains some elasticity during the cutting to the point that where the cutting and rebounding streams are not carefully confined, the particles can escape upwards into the cutting room. Once in the air they move at high speed, and bounce around the room, so that they can reach unanticipated places and can also be a hazard to folk doing the work.

Figure 1. Slot depths cut into granite by steel shot (left) and garnet (right)

The second problem relates to the cutting effectiveness. When cutting a brittle material the steel shot has a number of advantages, since the energy on impact is focused in the very small volume of the sphere in contact with the target. This improves the ability of the shot to generate and grow cracks in the impact zone and thereby improves the performance of the cut, over that of the mineral abrasives.

Figure 2. Relative performance of steel over garnet and sand in cutting dolomite, under otherwise similar conditions

However, when cutting ductile materials, such as metal, steel shot is not a good tool, since the focusing of the force means that the shot may get buried or just rebound from the target, without the tearing and plowing action that comes with the use of a more regular abrasive. One way to overcome the problem is to switch from a steel shot to steel grit, which is also available. The relative benefit can be illustrated by using the change from using glass spheres to using them after they have been broken into sharp fragments.

Figure 3. Effect of change in particle shape when using glass particles in cutting ductile composite material (after Faber and Oweinah).

This, by itself may not be a complete answer, since the process of making the grit makes it a little more vulnerable to abrasion and wear during the cutting process, but we have seen that it is possible to recycle most of the abrasive a number of times. However, because of the change in shape, it becomes a little more difficult to feed the abrasive into the cutting stream, and there have been occasions where the grit has bound up in the feed tube. This has, therefore, to be sized and the flow path designed, to ensure that this doesn’t happen.

Figure 4. Cuts made into tool steel using steel shot (left) and garnet (right)

The other change is to use a harder steel than normal. And here please note that there is a difference between the hardness of the steel and its toughness. As American Cutting Edge notes:

Hardness vs. Toughness: Generally as hardness increases, toughness decreases. Toughness is desirable when blades are heavily impacted, hardness when a blade is exposed to corrosive or abrasive materials.

Hardness is related to the amount of carbon in steel. Often the lower the carbon, the higher the toughness. Also, some steels do not perform at lower hardness as they were designed for use at higher hardness. . . . . . . . Hardness is a characteristic of a solid material expressing its resistance to permanent deformation. The Rockwell or Vickers hardness scales are most commonly used in the industrial blade industry.

Toughness on the other hand is the maximum amount of energy a material can absorb before fracturing, which is different than the amount of force that can be applied. Toughness tends to be small for brittle materials, because it is elastic and plastic deformations that allow materials to absorb large amounts of energy.

In general where the grit is being used to cut into other metals (which can include steels) the hardness of the cutting abrasive should be considerably higher than that of the target material.

Which comes to the third consideration, which is that steel abrasive can rust, and therefore, immediately after it has been recovered and washed, it should be effectively dried. This has proved to be more difficult to manage than originally anticipated, since, particularly where the particles are then stored for some time before re-use, any moisture present can create enough rust to “glue” the particles together. Which renders them effectively useless for further recycling and additional use.

So there are considerable pitfalls that can arise in making use of steel as a cutting abrasive, but where the jobs exist where it does effectively cut significantly better than the alternative (say in rock-cutting applications) and where the cutting zone can be shielded, and the particles rapidly recovered, dried and stored for relatively rapid recycling at an economic price, then it can be a productive way of reducing cost, while improving throughput. (And lest you think this is a new idea Gulf Oil did extensive work on abrasive jet drilling of oilwells starting in the mid-sixties, with some favorable results, but that is another story).

Waterjetting 28b - More on abrasive use

The costs of running a high pressure waterjet table divide into two parts, one that covers the basic costs of the system, whether it is running or not, such as building rent, while the second covers those costs that are a part of the actual work. Of the latter costs it is the cost of the abrasive that is often the most significant. This comes about in two ways, since the abrasive must first be purchased for use, and then, after it has been used it must be disposed of. Depending on the materials that were cut, this disposal cost can be significantly higher than the original purchase price. In some work carried out at the High Pressure Waterjet Lab (HPWL) at Missouri University of Science and Technology (MST) in the past we have seen disposal costs that were more than three times the cost of the original abrasive. And one should bear in mind that, as a research lab, the table was used much less than a comparable conventional table in an industrial cutting environment. But we also did not have a cutting operational budget, and so the cost of abrasive was something that we examined, to see if it could be reduced.

The first idea was that we would just recycle the abrasive. The particles of the target materials that were cut are generally much smaller than the abrasive particles themselves, and so it should be relatively easy to remove them from the mix. However, as we looked into the process in more detail, it was clear that it would not be quite as simple as it might, at first, appear. Marian Mazurkiewicz (retired) and Greg Galecki (who now runs the HPWL) carried out studies on the behavior of the particles as they moved through the mixing chamber and were accelerated down onto the target material. They found, as noted in an earlier post, that most of the abrasive was crushed to a smaller size when it passed through the cutting head, and a mix that started out with a particle size of 210 microns as it was fed into the system, was leaving the focusing tube with an average size of 140 microns.

Figure 1. Percentage of abrasive at different sizes after it has passed through a mixing chamber (and before it has hit the target). (After Galecki).

The reason that this is a concern is that, as the particles become smaller, so a point is reached where, depending on the target material, the abrasive no longer has sufficient energy to effectively cut into the target. When cutting into metals such as titanium and steel, our targets of choice in the study, this cut-off grade was at around 100 microns.

Figure 2. Effect of particle size on the cutting performance of an abrasive jet in cutting steel. (The tests were part of a factorial experiment and are thus averages over a number of different test runs at differing abrasive feed rates (AFR and pressures).

Roughly 25% of the mix in the example shown in figure 1 lies below 100 micron at it leaves the chamber. After impacting the target this value increases to more than 50%. Obviously recycling this fine material and re-using it in the cutting process is going to be less effective than removing it from the mix. Generally alluvial garnets will break up more rapidly than mined garnet, because of the structure of the abrasive particles, and thus the percentage that leave the focusing tube at the larger and more effective diameters are lower with the alluvial mix. The results were, we found, confirmed in the cutting results, with alluvial garnet producing a generally shallower depth of cut that would be achieved, other things being equal, in the cutting tests.

A quick word of explanation of the tests we ran, which are described in more detail here. The tests are run at a standard pressure and nozzle size, and at a constant traverse rate, with the depth that the jet cuts into a standard steel at a fixed speed measured over a 4-inch traverse length.

The results of the tests showed that, because of the particle crushing during the cutting process, the abrasive would have to be screened, and for most effective re-use only the larger fraction (on average less than 40%) should be recycled. The rest would be too fine for effective re-use in the operations we were developing. (Although finer abrasive has use in other applications, it would have to be screened and stored). It was interesting to note, and perhaps logical in retrospect, that once the particles had been used once and the larger ones separated out, then the percentage that survived and could be reused a second and third time increased significantly. This is mainly because those particles that had some form of weakness crack (either from weathering or from the mining process) were broken during the first impact, and the particles that survived did not have these cracks, and would therefore inherently be more prone to survive multiple times.

For our purpose, therefore, given that there was a high cost in purchasing the abrasive, and an even higher one in disposing of the contaminated material after cutting (because of the contamination by the target material) there was a potential economic advantage in recycling the abrasive. There were several ways in which the particles can be separated, but a simple screening process, if carried out properly, is quite time consuming, since the particles are required to “sit” on a vibrating screen for several minutes to ensure accurate separation, and this can be labor intensive if it is carried out as a batch process. We tried a number of different ways, including using a counter-flow fluid column that worked well for low feed rates, but the most efficient unit for one operation (we build virtually all of ours, and extensively modified them over time) may not be the best in other cases. (The one that survived the longest was a Wilfey table (though not this one).

In conventional AWJ cutting the abrasive has also to be dried before it can be re-used, and that can also add power and labor costs to the process. Thus, as with many choices that must be made when developing an efficient cutting operation, the best answer is to carry out a series of tests yourself, and run the numbers to decide whether, in the long run, recycling would, or would not, be an effective choice.

Tech Talk - A Gentle Cough!

When I last wrote about the global supply of oil, it was back in October, as the fall in oil prices was developing. Since then the price has continued to fall, with prices now below $60 a barrel. I was doubtful back then that the price would fall as far as it has, and remain cynical that it will remain down for very long. Since this seems to go against much current wisdom, let me explain why I remain pessimistic that the boost to the global economy from access to cheaper fuel will continue for any great length of time.

It depends on whose data you believe credible as to how much more oil is available than that currently in demand. When looking at the numbers in the past I used a number of roughly 1 mbd, but this is hard to realistically quantify. Why – well the problem comes with the regions of the Middle East and North Africa (MENA) where there are current conflicts. The ones of particular concern are Libya and Iraq, although the fluctuating state of exports from Iran cannot be neglected. When the Libyan conflict first impacted the export of oil from that country Saudi Arabia began increasing its production to offset the loss in Libyan exports.

There came a time in September when Libyan exports, which had fallen to around 300 kbd from a high of over 1.6 mbd, shot back up to around 900 kbd. The EIA has recently shown an inverse correlation between Libyan production and oil price:

Figure 1. Brent Oil Price and Libyan oil production (EIA )

Thus, when an additional 600 kbd suddenly appeared back in the marketplace, it is not surprising that it had an impact on prices. However while there was already some surplus in the market (from increased production in the US etc, as I will comment on below) the volume of the addition had a more significant impact on prices, and when KSA decided not to reduce production this led the market to assume that we had returned to plentiful sufficiency, and prices have continued to fall since.

However, this perception is already unraveling. Libyan conflict has continued to embroil their oil fields. The Sharara field, which produces 300 kbd closed in November as conflict overwhelmed it. At the moment two of the oil export terminals are threatened, and with them another 300 kbd of oil. But it is not possible, at this point, to predict what is going to happen in either location. There is little sign that the conflict is any closer to resolution, meaning the production will continue to be threatened into the foreseeable future. Sadly it it more likely that this will have negative impact on oil production, so that it might be wiser to assume lower rather than higher volumes coming from the country.

The situation is a little clearer and more optimistic in Iraq, where the pipeline through Kurdish territory has lessened the impact of the Islamic State take-over of a large swath of the country. The recent agreement between the Iraqi Federal Government (IFG) and the Kurdistan Regional Government (KRG) approved early this month is already raising questions over the volumes that the KRG will put onto the market. The agreement calls for sales of around 550 kbd, but there is an additional 100 kbd that is available, the status of which is unclear. The country is exporting, overall, around 2.51 mbd and the pipeline to Turkey is currently carrying 280 kbd, but is being boosted to carry 400 kbd, with an ultimate throughput of 700 kbd. Part of the problem in assessing the market for this, however, in the short term is that the Iraqi crude is often heavier and of relatively lower quality than the market average. This is currently causing some marketing problems, leading the IFG to lower prices in order to find a market. In neither case, however, is the current conflict likely to impact the production for export, and while it is difficult to anticipate much production above 3.5 mbd. (The December OPEC MOMR suggests that they are producing 3.36 mbd at the moment) we are unlikely to se any significant reduction in production going forward. The significant growth in global production to meet a still predicted rise in demand next year (albeit down slightly from previous estimates) will, therefore, not come from OPEC, who still anticipate that they will produce, on average 400 kbd less than they have this year. It is still expected that American production will continue to rise to meet expectations of increased global demand.

The problem, unfortunately, with that view, is that increases in US production are tied to output from fracked horizontal wells that are expensive to drill, and have a relatively short production life, with the majority of production coming in the first year of operation. Thus, in order to sustain production, more wells must be drilled each month to cover the loss in production from existing operations. The North Dakota Department of Mineral Resources projects that 225 or more drilling rigs are needed to sustain the growth of production from the state over the next three years (at which time it will plateau at around 1.5 mbd). Presently there are roughly 180 rigs operating, with the count falling by the week, as the rewards, at present, do not match the cost. The agency anticipates that the number will fall by an additional 40-50 rigs by the middle of next year. Well completions are also falling by the month, as the industry likely plans to wait out the current hiatus in prices. The impact of this on even short term production should not be discounted. There has already been a slight fall in production, rather than a gain, in October, and that will likely accelerate.

Without any gain in production, and in fact seeing the potential for a drop in US production over the next year, then the anticipated surplus between oil supply and demand will likely disappear. Remember that the MENA nations are seeing a growth in their internal demand for oil (in the KSA this has already passed 3 mbd) so that if they had no impetus to reduce production and exports in the face of falling prices, so they are unlikely to increase production when prices pick up. (They haven’t before).

When will this all happen? Well I got the size of the price fall wrong, so don’t hold me to the exact timing, but I would anticipate that when we see the start of the driving season next year, the oil market will tighten rather quickly. Following that (given the inertia in getting production back in the US) we will (as I have been expecting for a couple of years) see the global concern over supply start to be a significant factor in 2016.

Have a Happy Holiday!

Waterjetting 28a - Cutting rock on a table

In previous posts I have written about the use of lower pressure water (around 10,000 psi) as a way of cutting through rock. From the time that we first made a hole through nine-inches of granite while I was a graduate student some 50-odd years ago the way that we have recommended that rock be cut, in a mining situation, has been to use lower pressure and higher volumes of water. This is so that as many natural fractures around the individual grains and crystals can be developed at one time as possible. However the result of this is that the cut progresses along the grain boundaries of the infrastructure of the rock, that a roughly edged cut is made, rather than a smooth cut surface. In mining applications this isn’t necessarily a bad thing, but when cutting counter tops and other ornamental structures for marble and granite surfaces inside a building then a rough surface is definitely not often required.

So how can a relatively smooth surface finish be created along the cut? One step,that works with softer rocks (such as some pink granites) is just to increase the jet pressure, while at the same time reducing the volume of the stream flow. Once one reaches the ultra-high pressure regime (which is, for this article, considered above 35,000 psi) and with jet diameters on the order of 0.01 inches or less, the jet stream is more typically going to cut through a crystal within the rock than to just work on the cracks that lie at the edges of the crystal.

Unfortunately there are sufficient cracks and crystal boundaries within the rock that it is not possible to ensure that at some point as the jet cuts down through the rock, and along the desired path, that it won’t find a crack at a critical length and alignment that the crack will break out a larger chip. This is less likely to happen within the cut, since the confinement of the surrounding rock acts to reduce excessive crack growth, but can quite often occur at the rock surface, particularly where there has been some earlier damage that has left larger cracks within that surface. (This includes heat treatment).

That, however, is a specialized case, and in the more typical situation an increase in jet pressure to 50 ksi will not, by itself, produce the clean edge needed. Part of the reason for this comes from the striation planes within marbles, which can offer an easier path for the jet to penetrate, as the cuts get deeper, rather than having the hole continue forward along the jet axis. To overcome these problems it is easier, with the ubiquity of abrasive waterjet systems, to instead change to add an abrasive to the waterjet.

Dimension stone (the trade name for the decorative rocks such as marble and granite) is generally through cut with slab depths that are less than an inch-and-a-half thick, although greater depths can be specially prepared. Often the slabs are polished before they are finally cut to shape. We found that preferable, since when doing the final polish with successively finer grinding wheels (used for example in creating the Millennium Arch) the edge stress that can be generated by the wheels themselves can cause chipping along the edge of the work. This, in turn, either requires a regrind down to remove the chip, or some form of repair, which we found it difficult to make invisible given the complex structure of the granite. This is particularly true when relatively narrow ribs of material are being cut. As an example, consider the cartoonish mining figure that was made some years ago.

Figure 1. Toon miner carved from 3-inch thick granite.

The front and back surfaces were polished before the figure was cut from the slab, given the extreme fragility of the edges of the pick, for example, which failed under very little pressure in several samples before one survived.

One problem with this approach is that the edges of the cut, while relatively smooth, do not have the polished look that the flat surfaces have. Apart from making the cut relatively slowly, in order to remove as many striations along the cut path as possible, one answer has been to use a spray on the rock surface which then gives the impression of having a polished surface, and as long as the object is kept inside the coating will likely remain. (When we tried this with pieces that ended up outside weathering removed that coating within a short number of years).

The problem with hand polishing large flat surfaces is that it becomes very difficult to maintain a truly flat surface over the entire block, and while the surface may end up smooth and polished, it will likely have some small undulations within it. It is therefore more productive (and, we found, cheaper) to have large flat surfaces machine polished before they were cut. One example of this was the sign that we made for the State Geological Survey. It was made in two parts, the lower part was a Missouri Granite, which held an upper half, carved from Missouri Marble, which was cut to the shape of the state.

Figure 2. Sign cut for the State Geological Survey

The lower granite slab was inset into two vertical grooves that were cut into the supporting blocks. The granite slab was cut to shape on our cutting table, with the inset cut out to hold the “toe” of the state. Because the granite was first machine-polished the lettering was etched into the surface using a reduced pressure for the cutting jet, and removing a thin layer of the surface, which was replaced with the black fill material to highlight the letters.

Figure 3. The Agency name was etched into the granite slab.

When it came time to cut the shape of the state in the marble, the block was first trimmed at the top (to help it fit into the table). A piece of plywood was placed under the rock before cutting to prevent any rebounding abrasive from hitting the under side of the slab and removing the polish from the surface.

Figure 4. The first cut across the marble, showing the supporting plywood.

The rest of the state had a contour cut along each surface, and when these were completed the slab was ready for mounting.

Figure 5. The finished slab, showing the state outline.