In our latest well, CMR T2 cutoff at 33 ms is over-predicting movable saturation in a 0.3 mD Jurassic sandstone, while core Pc at 500 psi points closer to 6 ms. How are you tuning T2 cutoffs and choosing SDR vs Timur-Coates for k when paramagnetics skew the distribution, given the impact on net pay and flow potential?
@petrophys_mike I’d re-anchor T2cut to core by matching BVI to Swir at your 500‑psi Pc, but as a phi-dependent cutoff and after screening paramagnetics with a quick T1/T2 check (they’ll shove the spectrum left), then use Timur‑Coates for k with BVI‑calibrated C,m — one-size 33 ms is like using the same choke on every zone. What TE and temperature/salinity did you run on the CMR, and did you try this workflow: Petrowiki | OnePetro.
I’d bin the blanket “33 ms” and calibrate surface relaxivity from plugs against MICP (r35) so T2cut lands near about 6 ms at your 500‑psi Pc. For k, go Timur‑Coates when paramagnetics are active; FFI/BVI is steadier than SDR’s T2ML^2 once T1/T2 creeps >1.6. If you’ve got CMR dual‑TE, run a quick T1/T2 check and re-fit m,n to core k over a similar mineralogy window or you’ll overcall pay. @s_davies1990 have you seen TC stay stable around 0.3 mD?
, in 0.3 mD Jurassic rock the “33 ms” cut almost always inflates FFI when paramagnetics bite. Try reprocessing the CMR with dual TE to estimate internal gradients and pull the short‑T2 tail back, then set a phi‑dependent T2cut tied to your Pc match via r35 from MICP — this usually lands in single‑digit ms without overcalling pay. For k, I’ve had SDR outperform Timur‑Coates after that correction; any MDT/DST mobility you can use to choose the model, @zjensen405?
@amaya_t92’s point on skew is spot on — fix the physics first: apply Prammer “G‑correction” (or proxy from T1/T2) so T2ML isn’t shortened by internal gradients, then tie T2cut by matching BVI to Swir at your Pc point and run SDR on the corrected T2ML with a k cap from r35; do you have T1 or multi‑echo to back that up? That way the big‑hammer cutoff doesn’t smash your net pay.
Before picking SDR or a Coates‑style transform, let the flow test be the referee: match plug‑ and log‑derived k to MDT/mini‑DST mobilities, and tune the T2 partition so cumulative porosity above cut‑off predicts the same kh across tests. If you haven’t, run plug NMR at reservoir T on native brine to get ρ2 from r35, then temperature‑shift the log T2; paramagnetic skew often shrinks once the T correction and native brine are honored. @amaya_t92, do you also segment cut‑off by clay volume, using a higher bound‑water threshold where CEC is high?
I’d ditch the canned “33 ms” and back-calc surface relaxivity (ρ2) from plug NMR + MICP, then map a pore-size cutoff; in my last Jurassic tight sand that landed right near your “6 ms” at 500 psi. For k, Timur‑Coates beat SDR once BVI/FFI were re-partitioned with the new ρ2, especially where paramagnetics shorten T2; if OBM’s in play, a high T1/T2 ratio is a red flag. Do you have plug NMR to do a quick ρ2 fit, or want the Petrowiki recipe: Petrowiki | OnePetro?
Try a dual‑TE CMR to pin down internal gradient effects in‑situ, then force Dean‑Stark BVIr at stress to match the cumulative T2 curve — on a 0.3 mD Jurassic for us that shifted the movable/immobile threshold from about 30 ms to about 6 ms. For k, bin by FZI and let SDR handle clean units while a Coates variant fits clay‑rich zones, validated against MDT mobility — it’s like using a Phillips and a flathead, @t_graham76. Do you have Dean‑Stark and stress‑corrected MICP on the same plugs?
Use cuttings magnetic susceptibility to scale ρ2, push cutoff near ‘6 ms’; Timur‑Coates outperformed SDR here. Fe‑clays present?