Waterjetting 32d - Cutting with polymer in the water

In the recent past I have written about the use of polymers in high-pressure jets and that they can significantly improve jetting performance, with no additional changes in the power or pump and equipment used in the work. This is because of two different effects that the polymers have. Firstly they reduce turbulence in the flow from the pump to the nozzle, reducing pressure loss and increasing fluid flow, for the same pump power. As a practical consequence since the fluid flow will be greater for the same pump pressure, this will require that a larger orifice diameter be used to handle the greater, or – for the same flow rate and nozzle diameter, the pump can be operated at a lower pressure.

Figure 1. Comparison between a conventional jet and one containing the polymer additive marketed as SUPERWATER. (after Glenn Howells)

This is not immediately apparent, since the jet carrying the polymer appears smaller, but this is due to the second effect of the polymer, which is to tend to glue the water together, so that it is not dispersed as easily by the surrounding fluid – air in this case. In fact, for the same pump pressure and orifice diameter the lower jet will be operating at a higher pressure (since there is less pressure loss in the line) and there will be more water coming out of the polymer-supplied orifice at a higher velocity. But because it is not spreading into the air, it appears smaller.

This will improve the cohesion of the jet as it moves away from the jet, and as noted in an earlier post, this means that the jet will cut to a greater range from the jet, since it maintains the required critical pressure further.

The impact that the more coherent jet has on performance can be seen where the jet is used to cut into two different types of limestone, one oolitic and one crystalline.

Figure 3. Depths of cut of the polymer-containing jet (left) and plain waterjet, operating at the same pump pressure, nozzle diameter and standoff distance (Glenn Howells)

Note that at the distance where the normal waterjet has broken into droplets (as seen by the nature of the cut surface) and the jet has barely enough energy to remove the surface layer, where the jet contains polymer it retains the ability to cut.

Further, and this is more critical where cut quality is more important, the jet cut is much straighter and cleaner than the dispersed and wider normal cut. This can be seen where a different, more crystalline limestone has been cut closer to the nozzle.

Figure 4. Change in the cut shape to a narrower, deeper cut where polymer is added to the jet stream (rhs)in cutting limestone. (after Glenn Howells)

The benefit of the improved performance changes therefore with the distance of the target from the nozzle, with the more dramatic improvement being seen as the target gets further from the nozzle. Looking at the data from the original work that we did in Leeds, back in the early days of this study, this can perhaps be better realized through the use of a 3-D plot.

Figure 5. Improvement in cutting performance as a function of distance from the nozzle and jet pressure.

Note that, in these trials, the polymer improved the jet performance relatively more at lower pressures and greater standoff distance. Part of the reason for this (in hindsight close to 50-years after running the tests) is that when the jet was cutting at the lower pressures it was closer to the threshold pressure of the rock and this any drop in jet pressure had a more significant impact on cut depth than occurred at the higher pressures, where the gain was not relative to such a low benchmark.

For a number of years, until our research took us into fields where use of the polymer was precluded for several reasons, we routinely used a polymer (generally Superwater, marketed by Berkeley Chemical) rather than Polyox because it gave a relatively consistent and significant improvement in performance plus, being a liquid, it was relatively simple and inexpensive to buy a small metering unit (about the size of a small case) which would feed the polymer into the water supply line to the pump at the required concentration – typically 0.1 to 0.3%.

It tends to work better in improving abrasive jet cutting when it is used with a Direct Injection of Abrasive (DIAjet) or Abrasive Slurry Jet (ASJ) system than with conventional abrasive waterjet (AWJ) systems. The reason for this is that the AWJ system has the abrasive fed into the water stream at the mixing chamber just before the jet leaves the nozzle to strike the target. Within the mixing chamber the abrasive has to penetrate into the waterjet stream in order to acquire the jet velocity, and to distribute across the jet and give an even cut on the target.

Where the abrasive is feeding in from one side of the jet and the waterjet stream is more coherent, it becomes more difficult for the abrasive to penetrate the stream, and if the design is not adjusted accordingly, the cutting performance can be diminished, particularly relative to the gain that can be achieved where the combination is carried out effectively.

On the other hand with the ASJ systems the abrasive is mixed with the water far upstream of the nozzle and the are already thoroughly mixed together, so that the added cohesion of the jet will help to provide the acceleration that the particles need to reach close to the waterjet velocity, and achieve the improved cutting performance required.

Where an ultrahigh pressure jet does not contain abrasive, the polymer can be of benefit as a means of improving cut quality – as evidenced from this comparison in cutting a shoe sole pattern, both with and without the Superwater polymer.

Figure 6. Shoe sole cut comparison with and without Superwater. Note the smoother cut, with less fraying of the back side of the cut with the polymer. (after Glenn Howells)

Waterjetting 32c - more tests with polymers

In the last post on this topic I pointed out that one of early drivers to the use of long-chain polymers in water came from the reduction in friction that it provided to fluid flow through long pipes. In many instances this has been the driving force for the selling of the product, and in industries such as oil well drilling and fracking the reduction in friction down long relatively small diameter drilling pipe has been a significant selling argument.

The cohesion of the jet, once it leaves the nozzle, is a secondary consideration in overall economics, yet in some applications, such as the cleaning of down-hole completion screens, the ability of a polymer-laden waterjet to penetrate through the pressurized fluid in an oil well to reach and clean the screen has been the main reason that the market developed.

Figure 1. Improved jet power underwater when polymer is added (after Zublin)

One of the first steps to be addressed was the practical considerations as to how we got the polymer into fluid that made up the jet stream. The original polymer that I used was polyethylene oxide (Polyox) which was marketed in the form of small prills of chemical. The problem that we were then faced with is that, when these are just dumped into a container full of water, that the outer edge of each prill soaked up some water, became gel-like and adhesive, and stuck to the next particle, in a way that made a large collective lump that was very difficult to dissolve into the surrounding water flow. Even when the particles were fed in slowly into a fluid mixer the particles initially tended to concentrate in one layer of liquid, which only slowly dispersed into the main body of the fluid. That concentrated polymer has a number of interesting properties.

Figure 2. Lifting a thick concentration of polymer from a bucket by hand.

For example it can be thick enough that one can grab it with one's fingers and lift it that way out of the bucket, as the picture above shows, or it can cause a unique problem in a mixing tank.

The polymer can wind up around the mixing paddle shaft and work its way up the shaft until it hits the retaining screw at the top. It then piles up at this point until it reaches a critical mass, when a tendril can be thrown out of the tank, through the centrifugal force exerted through rotation of the paddle shaft. The tendril falling outside the tank falls to the floor, which is lower than the fluid in the tank, and thus the concentrated layer of polymer is drawn up the inside of the tank, over the side and down the outside of the tank since it is still attached to the escaping tendril. The result clearly showed that liquid could flow uphill, when pulled by the cohesion inherent in the high concentration polymer.

This, in turn, gives either a disadvantage (if you are using this in a factory) or an advantage to the use of the polymer. The reason comes from the fluid nickname – Slippery Water.” The addition of the polymer, while reducing friction in the pipe, also reduces it between a person’s shoe and the floor, and thus it becomes a hazard in the workplace, since it increases the risk of slipping. It has the impressive title Anti Traction Mobility Denial System . We used to call it Banana water, but that seems to have faded from use.

The need to reach the very low concentrations of polymer that are all that is necessary to enhance jet cutting required a better way of mixing, The recommended answer was to briefly suspend the particles in a suspension of isopropyl alcohol (swirling it in a cup worked well) and then dumping it into the tank in a way that ensured that the individual prills were distributed away from one another. And while this worked, it was somewhat cumbersome and worked well only when mixing up individual batches of water – useful in a laboratory but not so much in a factory that must operate steadily for a full shift.

A number of different chemical liquid additives, most particularly polyacrylamides and derivatives of guar gum, have been tested, with the original work (carried out with the help of Dr Jack Zakin) being carried out in special section of the Baxter Springs plant where we could photograph jets at one-millionth of a second in order to study their structure. To do that we set the system up so that the jet was back-lit, so that we could determine how solid the core jet was, and used a high-speed strobe to illuminate the jet for the short-time needed to freeze the jet motion, leaving the camera shutter open for that time. This meant that the room was totally dark, and since the tests were carried out in the middle of summer, it made for an interesting couple of weeks.

Figure 3. Improved cohesion of a 30,000 psi jet when polymer is added (lower picture) the jet range shown in the picture is about 8 inches.

We also ran a pressure transducer across the different jets, at different standoff distances, so that, for the most promising additives, we could measure the differences in impact pressure and jet cohesion as the transducer moved away from the nozzle. The results were reported in the Proceedings of the 3rd ISJCT with the different chemicals tested ranked according to their ability to improve jet cohesion and reduce jet spread.

One of the problems with some of the additives is that they are temperature sensitive, and the jet was coming from the nozzle at temperatures between 95 and 115 deg Fahrenheit (it was a hot summer and the water reservoir was not chilled). This was not recognized at the time, and it did have some impact on the performance of some of the chemicals, which also showed a tendency to rapidly age once mixed, due to the storage conditions. Nevertheless the results showed that while Polyox was the best compound, there were liquid alternatives that also were effective, and the technology has since switched to liquid additives of which I will have more to say next time.

Zublin, C.W., "Water Jet Cleaning Speeds - Theoretical Determinations," 2nd U.S. Water Jet Conference, Rolla, MO, May, 1983, pp. 159 - 166.
Zakin, J.L., Summers, D.A., The Effect of Visco-Elastic Additives on Jet Structure," paper A4, 3rd International Symposium on Jet Cutting Technology, Chicago, IL, May, 1976, pp. A4-47 - A4-66.

Me, This Week in Energy, and why Halliburton might refuse EPA

This afternoon I was an invited guest on “This Week in Energy” with Nikki Gordon-Bloomfield and Bob Tregelus. Among other things we talked about the fracking process that is being used to help produce the natural gas from shales such as the Marcellus and Haynesville. In the course of the discussion I was asked why of the nine fracking companies that EPA asked for their formulae, only Halliburton had refused the request. Bob pointed out that they were going to be subpoenaed and thus would have to give up the information anyway. I have discussed some of the problems of stimulating a well with hydraulic fracturing, both real and less so, on this site last March when the public perception of the technology began to change.

In the possible explanation I am going to give, you need to know that this is purely a supposition and the chemicals that I am going to mention are put forward out of my own head, as it were. I have no real clue as to why Halliburton are acting the way that they are, and am only building a hypothesis that might only through some slight possibility have any approximate relation to the truth.


In the evolution of the technology that has made production of the gas shales possible several different technologies had to be developed. The rock (which is actually a mudstone) has a very poor natural permeability. I.e. it is very difficult for fluid to flow through the rock, because the passage ways are very narrow and not very well connected. Thus the normal vertical wells would not produce very much oil or gas when drilled through the shale reservoir, certainly not enough to be profitable. The first beneficial development was, therefore, the ability to drill horizontally after the well had reached the reservoir depth. Once this was possible, then the length of the well that was exposed to the reservoir (which might be only 30 ft thick) would increase from that 30 ft to perhaps 10,000 ft. Since the amount of fluid flowing into the well is a function of the length of the exposed well in the rock, when the reservoir rock has a normal permeability this is enough to increase production significantly (as for example in the new wells in Saudi Arabia).

However when the rock has a very poor permeability even the long wells will only very slowly accumulate fluid from the surrounding rock, since there are no easy passages to the well through that rock. Thus the next benefit that was needed was the ability to crack the rock around the well. This is known as hydraulic fracturing or hydrofracking for short. In modern wells, by isolating and then pressurizing different segments of the well in turn, these cracks can be created (when the pressure inside the well exceeds the rock strength) at regular intervals (say 30 to 120 ft apart) along the length of the borehole.

The cracks are controlled in length (since if they go outside the reservoir all the fluid can drain away through the other end of the cracks, not to the well). But the problem is that once the crack is made, the pressure inside the well is lowered and the equipment moved to the next segment. Without any other changes as the pressure comes off the crack it will close back up, and there will not be much gain from the effort. So to keep the crack open what the industry calls a proppant, but you or I might just call it a carefully sized sand, is mixed with the fracking fluid before it is injected into the well.

As a result, when the cracks open in the rock, and the fracking fluid flows into the crack, the sand is carried with it, and is then trapped in the crack, holding it open after the pressure is lowered. A passage then exists for the gas or oil to travel to the well and production of most of the rock volume becomes possible.

Well that was when the development of the gas shale deposits began, however it had not been going on very long when it was noticed that the sand was not flowing easily into the fractures, and without enough sand being carried far enough back into the cracks, production wasn’t nearly as good as it should have been.

At this point another development was needed. This came about when an additional chemical – what is known as a long-chain polymer (typically a polyacrylamide) - was added to the fracking fluid. These fluids are known as Friction Reducing Agents (FRAs) because they tend to make water stick together a bit, and create extremely slippery surfaces when they coat them. By adding these FRAs to the fracking fluid, the crack walls became slipperier and the sand particles could thus travel deeper into the cracks, holding them open more effectively and increasing gas production. The fluids were given the generic name “slick water”, so that the current state-of-the-art is a horizontal well that has had a multi-fracture, slickwater-hydrofracking operation run on it.

But the problems of the wells are not over. As I noted in my post yesterday, the mudstones contain a significant amount of clays. And the problem when clays get wet is that they get softer and clay particles can break away from the wall of the fracture (slaking). Over the different gas shale deposits the problems are not consistent, since each shale is made of a different set of constituent rock types and clays. But overall the problem that is now being evidenced, as Art Berman has commented a number of times, is that the wells are losing production faster and earlier than predicted, so that they cannot meet the overall targets that make the well profitable. Instead of the well lasting perhaps a decade, they are losing perhaps 60% of the flow in the first year, and are no longer worth operating after maybe three years.

With all that as background, here is a hypothesis to explain Halliburton’s actions. It is quite possible that the well failures are due to the clay failure in the shale reducing the crack effectiveness. Clay content failure can do this, once the fracking fluid has cracked and wetted it, by a long term softening (which will allow the walls of the crack to fold around the proppant particles, and close the crack as the walls move in), or simply swelling into some of the crack space, with the same effect. Alternately the clay particles may slake and break away from the walls of the crack, and over time build up small dams along the crack path, again blocking the fluid flow through the crack – any one of these mechanisms explains the production falls that are being seen in the industry.

So lets say that Halliburton has realized the problem and, for merely the sake of a discussable solution, changes the polymer that they use from a pure polyacrylamide (PA) to include polyethylene oxide (PO). One thing that PO does at much lower concentrations than PA is that it stops the fracking fluid from wetting the shale, and interacting with the clay. Because it is (or at least was when we did this) much more expensive than PA there is not normally any reason to use PO in the fracking fluid.

But let us say that Halliburton have tried this, and it works. Because it is a step change in the process (in the same way as horizontal drilling; fracking; and slick water use were each, in turn) then the company selling the new idea has a tremendous commercial advantage. They can promise you that your well will stay in production long enough for you to make a profit, while the competition cannot.

The world of hydrofracking contractors is small and engineers move around, so that commercial advantage does not last very long, and word gets out as to how it was done. But that takes time, first to find out what is causing the problem, then what the answer is in general, and then what the answer is in detail. Each of those steps might take a competitor a year. That gives you three years of advantage, when you can charge higher rates, and possibly put some of that competition out of business.

The problem is that if the competition sees that you have put PO in your fluid, instead of PA then they can immediately go and look up what difference that makes to the fluid. Knowing that it stops wetting immediately gets them past stages one and two and cuts the term of your commercial advantage from three years to one.

Would you want to give that up if, by lawyering and all those fancy tricks they get up to in Washington you could get the time that you have to release the content postponed by at least a year? Likely not, and since dragging out the process can extend the period of your commercial advantage, the longer you can keep kicking the ball down the street the greater your advantage, and the more benefit.

And I re-iterate this is purely a hypothesis that I came up with, and I have no connections that would suggest that this has any connection to reality.