Amol,
Let me try to answer in two steps. First about temperature measurement and then about the absence of communication and risk.
You are right that both measurements are thermometers. With temperature you can do two things: look at anomalies where you've been injecting significant quantities of warmer (or colder) brine in the past, or try to detect effects of the flow while you inject. In theory you could see gradient anomalies when you have downflow-upflow (across your pathway), but in practice this is very hard to spot. Kreso's suggestion to pump hot/cold slugs if interesting, but the only info you'd get is "yes, you have a leak", unless you know the geometry of your leak pathway and the split in injected volume between target formation and pathway. Since the gain in information with a slug is minor, it's probably better to do a temperature log after a 24-48h shutdown - the longer the wait, the better. You can easily see a temperature anomaly of 5 degC or less where your warmer fluid has been flowing, and if you know your permeable zones mapped during drilling, you can also look closer at where the leak would have to be. I have seen a temperature anomaly from a frac job 3 years after injection.
So what if you don't see anything? If you were injecting in the US, EPA will say you don't have "significant" fluid movement and that your well has mechanical integrity. Remember: temperature is very sensitive to the leaked mass. You can try and quantify the "mass" with modelling, but it's not something you can achieve with Excel and a few hours' time. Clearly, if you don't see temperature effects, the risk of aquifer contamination is negligible.
What other risks you could have? Injected brine is very corrosive and cement protection in the annulus does not extend to oxygen. Unless you have [O2]<50 ppb or so, even a relatively small leak will end up attacking you casing from the outside. Luckily corrosion speed depends on brine velocity and volume, so sufficiently small leaks may never compromise your casing integrity.
Let's step back further: you want to estimate the leak rate to assess risk. Your configuration is ideal: the leak would go through cement and the ToC is below the previous shoe, so any leak in the B annulus is in direct communication with your wellhead. There are two types of pathways through cement: those that appeared during construction (channels and chimneys) and are large and dangerous; and microannuli that may have appeared at the start of injection and are a nuisance at most. Since you seem to have started worrying recently, it may be that you had isolation at the beginning. So forget about logging for one moment: you have to do a draw-down/build-up test in the B annulus and get the leak rate from there. If you can sustain "production", then do a long enough test to capture the steady-state rate. Do you have a log of the "questionable cement"? Even a CBL would do, but an ultrasonic tool would help you confirm which type of pathway gives you a pressure response, and what caused it. But remember: a cement log will not tell you if you have a leak. This you know from your annulus tests. The log will tell you where and why. My suggestion is that you forget about logs other than temperature profile for now: none of them can give you the rate directly and vertical velocity is not too useful if you don't know the pathway geometry. Rather do a good DD/BU+flow test and spend some resources in interpreting it well.
Best regards,
Matteo
------------------------------
Matteo Loizzo
Well integrity consultant
Berlin
Original Message:
Sent: Apr 23,2016 04:18 AM
From: Amol Agrawal
Subject: Tracer logging for identification of behind casing integrity issue in water disposal well
Kreso / Matteo,
Thanks for the response.
As I understand, temperature logging is similar to a 'thermometer' and will provide me the temperature of the fluid / medium around the probe/tool. It would be the same with DTS as well just that DTS will provide the temperature profile along the entire length? If I need to detect channels behind production casing with a probe/tool inside the tubing, I am not sure if it will read any temperature difference at all unless the leaking fluid is significantly hot or cold compared to injected fluid. The current injection water ranges in temperature from 60-70 degC at surface with reservoir at ~48 deg C. How much contrast in temperature would be required to get any detection of the leak and how we attain that significant contrast along the length of tubing/casing to detect leaking channels? For the absence of signals to be interpreted as 'no vertical communication' we need to be sure if it was even expected to show-up in temperature logging.
Regards,
Amol
------------------------------
Amol Agrawal
Production Technologist
Shell Gabon
London