10 Tips for Successful DFITs
Tips and tricks to make sure your DFITs yield accurate and useful data
If you’re reading this, you probably already know about the Diagnostic Fracture Injection Test, pre-frac injection test, or Minifrac. If not, I would highly recommend reading through the following paper as it is one of the easiest (when performed correctly), most cost-effective ways to get some very valuable reservoir parameters (reservoir pressure or p*, closure stress/pressure, and permeability to name a few). The test itself involves pumping a small volume of fluid (5 -10 to upwards of 40–50 bbls) at a constant rate (ranging from 2– 20 bpm with the average normally around 5 bpm) into an isolated zone to create a fracture in the reservoir, then shutting the well in and monitoring the pressure leak-off over time.
While the tests themselves are not all exactly the same, the execution of the tests are very similar so we will treat them as the same test for the purpose of this post.
Today, I will discuss 10 tips and best practices that will result in higher quality data and more confident analysis.
1. Failure to Communicate
It is impossible to over-communicate with your vendors, field personnel, consultants, and other stakeholders involved in the test.
Between your field staff, consultant/company man, the pump company, and the gauge company, there’s a lot of people involved in such a simple operation. Miscommunication is where most issues occur. It is imperative that everyone be on the same page and have the same understanding of what you will be doing. We’ve broken these down into the following categories:
Selection of Equipment
Who will be providing the gauge? Is the gauge thermally compensated? Who will be pumping the job? Who (and how) will be monitoring the injection rate? These are just a few questions to consider and keep in mind when communicating the details about the job.
Installation of Equipment
When will you be rigging up? Where will the gauge be rigged up? Will you be monitoring an inline flow meter (the answer should always be YES — more on that later) and if so, whose meter will it be (gauge co. or the pump co.)? Will you need a stinger (wellhead isolation tool) to pump this job, if so — make sure that you plan on keeping that stinger on the well for the duration of the monitoring period as removing it will cause an interruption in the test and result in bad data.
Job Execution
Who will be calling breakdown, max rate, max pressure? Does the pump company know the importance of shutting in the wellhead post-injection before they start rigging down? Below is an example of what your test should look like (it’s important to note that you do not have to do a step down test at the end of DFITs, but it’s shown in this example) as well as the segments of the test from the paper SPE 169539-MS, Barree, Miskimins, and Gilbert.
There’s a number of things that end up going wrong during the actual job that cause bad test. Make sure you discuss these things with the team ahead of time so that they aren’t scrambling to figure out what to do while you’re pumping.
Some key things to remember:
- You need a steady rate to accurately observe a breakdown pressure — even if it’s not the exact rate that the procedure recommends! Best practice is to get your pumps up to your desired rate quickly (as long as it’s within the confines of your max pressure) and hold them there. If the design calls for 10 bpm but you can only reliably/confidently get to 8 bpm, that’s ok just go with 8 bpm. Trying to mess with your rate while you’re trying to see breakdown will cause issues.
- Make sure to reset your total volume parameter before you start injection. This isn’t a huge issue as long as your analysis software allows you to calculate from start of injection line or delete previous rates. There’s some different theories about whether you start calculating this volume at the beginning of the job or once you see breakdown pressure but that is outside the scope of what I am going to cover today. The basis of the test is that you want to inject a known volume of fluid and that volume of fluid is used to calculate your important reservoir parameters. If that number is wrong, you’ll get incorrect reservoir parameters.
- ALWAYS use an inline flow meter to monitor your injection rates and volumes. We have seen plenty of examples where a pump company is using pump strokes to monitor rate and it is very different than what we are reading through our inline flow meter. Inaccurate volume/rate data will lead to inaccurate analysis results.
After Job Procedure
This is another area where the opportunity to mess up the test is high. Everyone has let their guard down and is ready to rig down and head home.
One of the most common issues is the failure to isolate the wellhead before the crew begins rigging down the iron and equipment. You MUST shut the well in before you start rigging down. It is also important to make sure any valves to the surface gauge are left open and untouched. The service company needs to be able to RDMO without any disturbance to the wellhead since it’s a pressure sensitive test. Any change in well pressure after the injection will result in a questionable dataset at best. Make sure you clearly and directly communicate to the pump company that they are not allowed to touch the well until the consultant/engineer/company man gives the all clear that the well is shut-in and isolated.
2. Keep a Clean House
Make sure to clean and test all your iron and equipment before you start your test!
This sounds simple but it’s yet another common way these tests end up failing. You have no idea where the pump equipment came from or what they were pumping before they go out on location. This presents the opportunity for outside debris to cause issues on your equipment and cause improper readings. Make sure to have your pump company flush their lines thoroughly so that any physical debris has been flushed. Some examples of what I mean can be seen below.
It’s hard to tell but there is a blue bottle cap in the bottom right corner of that flow meter. This bottle cap resulted in no rate data to be acquired.
As you can see in this example there’s a rock that is blocking the flow meter from rotating, also resulting in no rate data to be acquired.
3. Avoid Conflict
Make sure that your procedures do NOT conflict with the actual rigup and installation of the job.
This can range from recommended injection rates and pressures to the equipment that is required for the job.
Below is an example where the operator had to use a stinger (wellhead isolation device) to pump their test. While this is fine, they did not consider the fact that they had communicated to their field staff to remove the stinger once the test was completed. Since you must monitor the pressure from the stinger itself, this resulted in a gap in the data (when they removed the stinger and moved the gauge back onto the wellhead) and thus a bad test. It is generally best practice and good risk mitigation to just leave the wellhead alone once the injection is done. DO NOT REMOVE OR RELOCATE THE GAUGE after you’ve started or completed the injection. Any chance in pressure will result in a bad test data.
Another example (below) shows how important it is to have a plan/procedure for when equipment fails. The procedure for this job called for injection to be pumped at 10 bpm. In this data set you can see early on, that one of the pumps tripped out, resulting in a drop in rate. They tried to bring the pump back online but it kept tripping out. This crew was prepared for this scenario, so they continued injecting but walked the rate down to 4 bpm which reduced the pressure enough to stop their pumps from tripping out. This allowed them to save the test and allowed for them to see a clear breakdown. The recommended injection rate does not always have to be met, you just need to ensure that enough rate is supplied to create a clear breakdown resulting in the creation of a fracture.
It is critical that everyone involved understands what to do in case of equipment failure or unexpected events (pumps tripping out) to ensure the test goes smoothly.
4. The Weather Outside is…IMPORTANT
Always acquire and consult the temperature data!
When in doubt, always check the temperature data first. There’s a number of things that can cause issues during a test but one of the easiest things to use to identify/eliminate them is the temperature data. Below is the late period fall-off data from a DFIT that has a number of (seemingly random) large pressure spikes.
Let’s take a look at the pressure data with the temperature data on the Y2 axis to see what might be happening here.
As you can see in the orange circles (highlighted in bright blue) those pressure spikes line up perfectly with temperature drops below 30 degrees fahrenheit (green dashed line). These pressure drops resulted in freezing in the wellhead/fittings, causing ice to form and the pressure to increase. When the temperature warmed back up, the ice melted and the pressure returned to the baseline fall-off. This won’t ruin your test but it can drive you crazy trying to determine what caused it. To help with your process of elimination when there are issues (let’s face it, in the oilfield, there’s always issues) always check the temperature data first!
5. Time to Understand
A procedure is only as good as it is understood.
As engineers, we are very driven by processes and procedures but we have to make sure that those are understood thoroughly at all levels. In this next example, the operator directed the field personnel to pump methanol into the well to mitigate potential freezing issues (similar to what we just saw in the previous plot). This should have been done during the test, but that was not communicated to the field personnel. As you can see in the green circle, they pumped the DFIT and then shortly after they shut-down, they opened the well back up and injected methanol (as they were instructed). This resulted in a bad test and a waste of everyone’s time and money.
As you can see, it’s very easy for things to go wrong if procedures aren’t fully understood or clearly communicated.
6. You Get What You Pay For
Data quality can be a very big factor in these tests yet it is often overlooked in an effort to reduce cost.
As an industry, we live in a boom-bust cycle and the pressure to cut costs is (more often than not) a factor in our decisions. I’m not advocating that you go out and buy the most expensive thing that you can find for these tests, but I am saying that not all gauges and equipment are created equal.
Make sure to educate yourself as much as you can on your supplier’s equipment type, accuracy, and resolution. It’s not a fun topic and it can be confusing at times, but when you’re trying to get accurate reservoir parameters, it’s important. Things like the type of gauge (strain-wire, silicon-sapphire, quartz), whether it’s thermally compensated, the accuracy and resolution (full-scale or reading) are all things that you need to consider when performing a DFIT. Check out the links I’ve embedded if you want to learn more about what all this means.
Because DFIT analysis is very reliant on pressure derivative and injection volume, it is critical to use high quality monitoring equipment. As you can see in the example below, there’s a reason the better quality data costs more — it’s more accurate! Here, we have a strain-wire type pressure transducer and an uncalibrated turbine flow meter monitoring the same data as a high quality, thermally compensated silicon-saphire pressure and a properly sized and calibrated flow meter. The differences are apparent with the naked eye and will directly affect your ISIP, closure pressure, reservoir pressure, and permeability outputs.
One area where it’s easy to cut some cost is to acquire the injection rate data from the pump company using pump strokes instead of paying for an inline flow meter. It’ll be fine, they’re both measuring rate data, right? Survey says….. WRONG. Below you can see a comparison of the rate data from an inline flow meter and the pump company’s rate which is being calculated based off the strokes of each pump. As you can see, they are not the same. This is yet another way to increase the reliability and confidence in your tests. ALWAYS MONITOR RATE USING A FLOW METER.
Acquiring actual (flow meter) injection rate is critical since injected volume is one of the most important inputs in DFIT analysis. Failure to have accurate rate data will result in incorrect permeability numbers.
7. Let’s Get Physical
Consider how physical factors affect your data and mitigate them as much as you can.
I highlighted the importance of having temperature data earlier due to temperature cycle at the surface and the freezing effects it may have on your data. Another area where temperature can affect your data is when temperature swings drastically throughout the duration of the day. This is one of the reasons it’s important to have a thermally (temperature) compensated gauge as it will help correct the pressure readings based on the temperature data. Even with thermal compensation, there are situations where you can still see temperature effects on the pressure data. This typically occurs when you see daily temperature swings over 30–40 degrees F in a short time frame (a day). In the plot below you can see how the large temperature swings in (blue) cause “noise” in the pressure data (red).
Some people would argue that this is a reason to run downhole gauges as this thermal effect is limited to the surface. However, this is not the case. As you can see in the dataset below (same as above but with downhole pressure), both surface and bottom-hole gauge data see the thermal effects. When you have temperature swings this large, the actual fluid in the wellbore experiences thermal expansion/contraction resulting in “noisy” pressure data. This happens whether the temperatures are extremely hot or extremely cold.
Unfortunately, there’s not a lot we can do to mitigate these effects as the wellbore itself is acting as a giant conductor — heating and cooling with the temperature of the earth. However, if you see something like this in your data, save yourself some time and start troubleshooting by looking at the temperature data.
8. Be a Nosy Neighbor
Be aware of concurrent or nearby operations.
The trend over the last decade has been more, more, more. More wells, more proppant, more fluid. This is something that must be kept in mind when planning and executing your tests as your well is just a single point in a much larger system. Nearby operations and offset activity can directly show up in your data and ruin your test.
In the example below we have pressure from a gauge monitoring a DFIT (red) on the left axis and the frac treatment pressure (blue) from a nearby well on the right axis. As you can see, the DFIT pressure is no longer declining because it has been interrupted by frac hits/interference from the offset well. We were lucky enough to be monitoring both of these jobs but in most cases, you won’t be monitoring both data sets and you will be extremely confused as to how your pressure data is now responding. We’ve even seen interference from a nearby well that was being drilled out after the frac.
Being cognisant of your nearby operations and planning your tests accordingly will go a long way in increasing the number of good data sets that you collect.
9. Trust but Verify
Always verify before moving to the next step in the procedure.
Make sure to communicate to all the field personnel how critical it is to verify everything is as they suspect it is before the move on to the next step in the procedure. Assumptions are the cause of a lot of failed tests but fortunately they’re easy to mitigate with enough communication. In the example below, the pump company assumed that the well had been shut-in and isolated, so they started rigging their iron down. As you can see from the data, this was not the case and they started bleeding pressure off the well, resulting in a failed test.
As I’ve mentioned before, once the pressure has been interrupted, the test is no longer reliable or accurate. Always have your guys in the field verify multiple times that every task has been completed before moving on to the next one.
10. MONITOR YOUR DATA
Stream your data and check it often.
At one point, this technology was rare and expensive. As with all things tech, this is not longer the case. These days, most reputable companies have the ability to stream or transmit their data in real-time (and if they don’t give us a call ;) ) for a nominal fee. I understand that this feature may be seen (in some instances) as a nice-to-have instead of a necessity. But in my experience the value greatly outweighs the small fee. As I’ve shown previously, there’s a large number of ways to mess up these tests and if you’re not on-site or don’t have access to the data, how are you going to know? Further, there’s a lot of instances where the actual injection and shut-down go perfectly but hours, days, weeks later some external event occurs that ruins the remaining data of your test. If you’re not monitoring it, how will you know? In this instance, you’d end up saving money by paying for the data stream because you’d be able to end your test earlier than planned.
Below, you can see a well executed DFIT injection that goes on vacuum after just 20 minutes. Had the operator not been monitoring this data, they would have left this gauge on the well for another 20 days — capturing useless data and wasting valuable time and money.
Wells go on vacuum, they get opened by pumpers, they get interfered with by offset stimulations — the list is long and there’s only so much you can do to mitigate for these things but if you’re monitoring your data, it won’t matter.
The other big benefit of real-time monitoring is the ability for real-time Analysis.
If you have access to your data, you can pull the data into your favorite analysis tool at any time and see where you’re at in your test. We have seen customers reduce their DFIT time and costs by over 50% simply by having the ability to update their analysis. Have I seen closure? What flow regime am I in? Do I have enough data that I’m confident in my analysis? These are all risks that you can mitigate by simply monitoring your data.
Conclusion
I hope these 10 habits/tips help you perform better DFITs and increase data quality and confidence in your resulting analysis. It is, by far, one of the best bang-for-your-buck tests that can be performed in the oilfield and I hope this post allows you to be more successful and acquire better data.
If you liked this post, please don’t forget to recommend and share it. Until next time…
— John Kalfayan
