Flathead Lakers v. DNRC

Majority Jim Rice

05/16/2023 DA 21-0535 Case Number: DA 21-0535 IN THE SUPREME COURT OF THE STATE OF MONTANA 2023 MT 85 FLATHEAD LAKERS INC., a Montana Non-profit Public benefit corporation, AMY J. WALLER, STEVEN F. MOORE, CYNTHIA S. EDSTROM, ADELE ZIMMERMAN, MARTIN FULSAAS, GAIL A. WATSON-FULSAAS, LAUREL FULLERTON, ALAN COIT, DEIRDRE COIT, and FRANK M. WOODS, Petitioners, Appellees, and Cross-Appellants, WATER FOR FLATHEAD’S FUTURE, Intervenor and Appellee, v. MONTANA DEPARTMENT OF NATURAL RESOURCES AND CONSERVATION, Respondent and Appellant, and MONTANA ARTESIAN WATER COMPANY, Respondent, Appellant, and Cross-Appellee. APPEAL FROM: District Court of the First Judicial District, In and For the County of Lewis and Clark, Cause No. CDV-2018-135 Honorable Kathy Seeley, Presiding Judge COUNSEL OF RECORD: For Appellants: Brian C. Bramblett, Chief Legal Counsel, Montana Department of Natural Resources and Conservation, Helena, Montana  Rick C. Tappan, Tappan Law Firm, PLLC, Helena, Montana (for Montana Artesian Water Company) For Appellees: John J. Ferguson, Graham J. Coppes, Emily F Wilmot, Ferguson Law Office, PLLC, Missoula, Montana David K.W. Wilson, Jr, Morrison Sherwood Wilson & Deola, PLLP, Helena, Montana Roger M. Sullivan, McGarvey Law, Kalispell, Montana Submitted on Briefs: November 30, 2022 Decided: May 16, 2023 Filed: ' ,-&-•-if __________________________________________ Clerk 2  Justice Jim Rice delivered the Opinion of the Court. ¶1 Defendants Montana Department of Natural Resources and Conservation (DNRC) and Montana Artesian Water Company (Artesian) (together, Appellants) appeal the Order on Remand entered by the First Judicial District Court, Lewis and Clark County. The District Court vacated the DNRC’s Final Order issued in In the Matter of Artesian’s Application for Beneficial Water Use Permit No. 76LJ-30102978, which had granted the permit, and remanded the matter for further consideration, on the grounds that deficiencies in Artesian’s application were material to consideration of physical availability of water, and that DNRC relied improperly on an internal agency memorandum (Memo) in analyzing legal availability, leading to a failure to sufficiently analyze the potential impact of Artesian’s proposed groundwater appropriation. Flathead Lakers Inc. and Water for Flathead’s Future (Objectors) cross appeal from the deemed denial of their motion for attorney fees. ¶2 Accordingly, we address the following issues: 1. Did the District Court err by determining DNRC erroneously granted Artesian’s application for a beneficial use permit? 2. Did the District Court err by denying Objectors’ motion for attorney fees? ¶3 We affirm in part, reverse in part, and remand for further proceedings. FACTUAL AND PROCEDURAL BACKGROUND ¶4 This matter returns to the Court a second time. In Flathead Lakers Inc. v. Mont. Dep’t of Natural Res. & Conservation, 2020 MT 132, 400 Mont. 170, 464 P.3d 396 (Flathead Lakers I), we determined the District Court erred by holding the permit issued 3  to Artesian was invalid because DNRC failed to require compliance with regulations regarding Artesian’s Application. We reasoned that the Application, despite its deficiencies, was rendered “correct and complete” as a matter of law upon the passage of 180 days after the filing of the Application without DNRC’s notification of defects in the Application, pursuant to § 85-2-302(5), MCA. Flathead Lakers I, ¶¶ 10, 19, 21. However, we explained that “[a] correct and complete application does not mean that the permit will be granted; the applicant still must show by a preponderance of the evidence that the [§ 85-2-311, MCA] criteria are met.” Flathead Lakers I, ¶ 18.1 We foreshadowed the merits challenge to the permit’s validity now before the Court: “The issue remains whether DNRC’s consideration of the application without the additional aquifer testing was arbitrary and capricious and whether its evaluation was clearly erroneous in light of the record.” Flathead Lakers I, ¶ 20. On remand, the District Court concluded they were. Thus, the essence of this appeal is whether the deficiencies in Artesian’s Application, DNRC’s processing of the Application, including employment of the Memo, and/or other problems undermined the validity of the agency’s analysis and determination to grant the permit. Application requirements and testing ¶5 Artesian owns land in the Creston area of Flathead County, on which it seeks to operate a water bottling facility. Artesian intends to pump 710.53 acre-feet of water 1 Admin. R. M. 36.12.1601(4) explains: “Providing correct and complete information is not the same as proving the statutory criteria. The department can only grant an application if the criteria for issuance of a permit or change application is proven.” 4  annually2 from a 222-foot deep well3 located on the property. In accordance with the Montana Water Use Act (MWUA), Artesian submitted its Application for Beneficial Water Use Permit No. 76LJ-30102978 (Application) to the Kalispell Water Resources Office of the Department of Natural Resources and Conservation (DNRC) in June 2015. ¶6 As part of the Application, Artesian conducted an aquifer test to assess, as stated in the agency’s Aquifer Test Report, “drawdown for use in evaluating adequacy of diversion and adverse effect of the applicant’s request.”4 The data from the test was entered into Form 633, the agency form providing the regulatory testing requirements, and entitled “Aquifer Test Data Form.”5 Prior to the testing, for reasons that are unclear, Artesian, upon finishing the drilling of the well on December 18, 2014, allowed the well to be continuously 2 588.08 acre-feet is apportioned for bottling and considered 100% consumptive, while the remaining portion designated for cleaning and other operational functions is estimated to be 10% consumptive. Up to 12.28 acre-feet of the bottling water is to be seasonally apportioned for geothermal heating. 3 The well is screened to withdraw water from a depth of between 203 and 221 feet. The static water level in the well is -28.8 feet, meaning that the water would rise 28.8 feet above ground if the well was of sufficient height. 4 “Drawdown” is the lowering of the water table. As groundwater is pumped from a well to the surface, the groundwater level, or water table, decreases. The difference between the water level before pumping and the new equilibrium level during pumping is the measured drawdown. The pumping of a well also results in a cone of depression wherein the local level of the water table is lowered in a cone shape centered around the well, as drawdown at the well causes surrounding water to redirect to the area of lower groundwater equilibrium. 5 Admin. R. M. 36.12.1703 requires applicants to “follow aquifer testing requirements and provide to the department, at minimum, information and data in conformance with Admin. R. M. 36.12.121.” In turn, Admin. R. M. 36.12.121 lists “minimum information that must be submitted with applications” and “[m]inimum testing procedures,” which are all to be reported in Form 633. Thus, the regulations incorporate completion of Form 633 as the mechanism for compliance with the regulations. 5  pumped thereafter at a rate of 175 gallons per minute (gpm) for almost three months until the required aquifer test was initiated on March 9, 2015. The aquifer test itself consisted of a 72-hour pumping period during which an average of 455 gpm was pumped from Artesian’s well. During the test, discharge rates were to be measured hourly,6 but in two separate 12-hour periods, none of the hourly measurements were recorded. The first omission occurred between the fourth and seventeenth hours of testing.7 Further, the groundwater levels at the well were not measured during the testing period at all. ¶7 The aquifer test also required background water level monitoring before the testing, and recovery monitoring afterwards, for both Artesian’s well and separate observation wells.8 Two existing wells, the Nickol and Koch wells, were selected as observation wells, and no new observation wells were drilled.9 On Form 633, the Koch well is recorded as 380 foot deep, while the depth of the Nickol well is marked as “90?” feet. Contrary to 6 Admin. R. M. 36.12.121(3)(c) requires that “[d]ischarge rate must be measured with a reliable measuring device and recorded with clock time according to the schedule on Form 633.” In turn, Form 633 requires multiple entries per hour for the first three hours, and then, at minimum, hourly measurements through the first twenty-four hours, and measurements every three hours thereafter. 7 Although Artesian failed to record the required hourly discharge measurements and therefore could not fully track fluctuations in the discharge rate, it used a flow totalizer from which it deduced the well’s average flow rate over the course of the test. 8 Admin. R. M. 36.12.121(3)(j) provides: “Groundwater levels in the production well and observation well(s) must be monitored at frequent intervals for at least two days prior to beginning the aquifer test to evaluate background water-level trends.” 9 Admin. R. M. 36.12.121(3)(h) states: “One or more observation wells must be completed in the same water-bearing zone(s) or aquifer as the proposed production well and close enough to the production well so that drawdown is measurable and far enough that well hydraulics do not affect the observation well. If existing wells are monitored they must not be pumped, or if pumped should be monitored at a frequency necessary to separate the effects of the pumping.” 6  Admin. R. M. 36.12.121(2)(c), the diameters and depths of perforation for both wells were omitted from Form 633.10 Over several days prior to the test, while Artesian’s well was pumping out 175 gpm, the background water levels were measured at both observation wells and Artesian’s well. Several of Objectors’ experts opined, at the later hearing, that the prolonged pumping of Artesian’s well ahead of the aquifer test, along with its continued pumping during the measurement of background water levels, would cause significant inaccuracy in any measured levels because the aquifer conditions around the wells were not static. ¶8 During the aquifer test, drawdown was measured at both observation wells. Both wells experienced drawdown, but not in excess of their artesian water levels.11 The Koch well water level was observed to drop and recover in a manner consistent with expectations for a well located within the same aquifer as the production well, here, the Deep Aquifer.12 However, the Nickol well responded and recovered significantly slower, evidencing a potential aquitard13 between the wells. Contrary to expectations for a nearby well, 10 Admin. R. M. 36.12.121(2)(c) requires the inclusion in Form 633 of “distances between the pumping well and the observation well, and depths, dimensions, and perforated intervals of each well as specified on Form No. 633.” 11 The artesian pressure at both wells allowed water to naturally flow to heights several feet above ground. 12 The deep alluvial aquifer in the Flathead Valley is known as the “Deep Aquifer.” Montana Bureau of Mines and Geology has estimated the Deep Aquifer to have an unappropriated water influx of 190,000 acre-feet per year. 13 An aquitard is a layer of semi-impervious rock or soil that restricts the passage of water through it. This layer is also be referred to as a “confining unit.” 7  drawdown at the Nickol well had not yet stabilized after 72 hours of pumping during the aquifer test, and did not fully recover for another 72 hours afterwards. Because of this delayed behavior, several experts opined that the Nickol well is located outside the Deep Aquifer.14 Agency processing of Application ¶9 Upon submission of the Application to DNRC, Nathaniel Ward, a DNRC water resource specialist with the Kalispell Regional Office, was tasked with reviewing the completeness of the Application. On August 28, 2015, Ward issued a deficiency letter notifying Artesian that additional information was required regarding Artesian’s means of diversion, means of putting the water to beneficial use, calculation of volume requested, and the authority of the Application’s signatory. Despite the above-referenced omissions of data required by regulations to be submitted on Form 633, the deficiency letter did not request the missing data. It would be later learned that neither Ward, nor Kathy Olsen, the regional manager of DNRC’s Kalispell office, ever reviewed Artesian’s Form 633 for compliance. ¶10 Instead, and while Artesian’s response to the deficiency letter was pending, Ward sent the hydrology components of the Application, including Form 633, to Attila Folnagy, a DNRC groundwater hydrologist based in Helena, for review. During his review, Folnagy noticed the data omissions from Form 633, but did not view them to be significant enough 14 This would disqualify the Nickol well as an observation well, as noted by several experts. See Footnote #9. 8  to notify Ward or anyone else of the omissions, nor inquire whether the agency had granted a variance from the data requirements. Admin. R. M. 36.12.1701(4) authorizes exceptions to the testing requirements if the applicant is granted a variance by DNRC. According to Admin. R. M. 36.12.121(1)(b), an applicant can request from the appropriate regional office manager a “variance from testing requirements” to allow the applicant to forego specified testing requirements. Rather, Folnagy deemed the supplied data to be adequate and used it for preparation of the Aquifer Test Report and the Depletion Report. In the Depletion Report, Folnagy incorrectly documented the initial date on which Artesian’s well began pumping; rather than the actual date of December 18, 2014, Folnagy listed the pumping as beginning on February 18, 2015.15 As noted, the purpose of these reports is to assess “drawdown for use in evaluating adequacy of diversion and adverse effect” of the proposed use. ¶11 In the Aquifer Test Report, Folnagy explained his methodology in analyzing drawdown. Folnagy employed the Neuman-Witherspoon Solution16 (the Solution) “to best simulate aquifer drawdown in a confined two-aquifer system” because of Artesian’s well’s placement in the Deep Aquifer, the existence of the shallow aquifer, and the understood leaky nature of the Deep Aquifer in that area. Folnagy inputted the data provided on Form 15 During the contested case hearing, Folnagy testified that he was unaware that Artesian’s well had begun pumping on December 18, 2014. 16 The Newman-Witherspoon Solution is a groundwater model created by Shlomo Neuman and Paul Witherspoon in 1969 to model the behavior of groundwater in a two-aquifer system in which the pumped aquifer is confined and leaky. 9  633 regarding Artesian’s well and the observation wells into the Solution to model the physical availability of water in the area and identify any potential adverse effects of Artesian’s pumping. Folnagy concluded therefrom that no impacted wells in the area of greater than 100 feet of depth would experience drawdown below their typical water column levels. Then, in his Depletion Report, Folnagy listed only the Flathead River and Flathead Lake as “Potentially Affected Surface Waters” because those water bodies “are identified as potentially affected surface waters in the January 10, 2011, memo regarding legal availability of groundwater in the Deep Aquifer.” (Emphasis added.) No other potentially affected surface water sources were identified in the report. ¶12 The January 10, 2011 memo (Memo) referenced by Folnagy is an internal document written by two DNRC hydrologists, including Russell Levens, DNRC’s supervising hydrologist in the Water Management Bureau, which provides a unique framework for analyzing legal availability for groundwater withdrawals from the Deep Aquifer. The Memo, consisting of one paragraph on a single page, has never been formally adopted as a regulation, but is utilized as an agency policy. According to the Memo, applications for withdrawals from the Deep Aquifer require surface water legal availability analysis to be conducted only for the Flathead River and Flathead Lake because “groundwater levels in the deep aquifer are effectively controlled by the surface water levels in the Flathead River and Flathead Lake,” due to “the upward gradient of groundwater in the valley.” The Memo cautions, however, that “[l]ocal areas of the deep alluvial aquifer may be hydraulically connected to other surface waters or reaches of the Flathead River. In those cases, applicants need to evaluate availability on those sources.” (Emphasis and underlining 10  added.) Thus, while recognizing that there “may be” other surface waters connected to the Deep Aquifer, the Memo, as written, places the burden on the applicant to identify and demonstrate legal availability in those connected surface waters for a proposed new use. Under the Memo, barring an applicant’s demonstration of other hydraulically connected surface waters, the only waters analyzed for legal availability by the DNRC are the Flathead River and Flathead Lake. ¶13 Testimony of DNRC’s personnel at the contested case hearing shed light on the evolving application of the Memo by the agency. According to Levens, at the time of the Memo’s creation, applicants bore the burden of identifying any potentially impacted surface waters, hence the phrasing of the Memo’s final sentence that “applicants need to evaluate legal availability of [potentially hydraulically connected surface waters].” Thus, when analyzing potentially affected surface water sources for legal availability, DNRC, in Levens’ words, “reviewed the work done by applicants. . . .” However, around 2012, as part of a larger effort to streamline the application process, Tim Davis, division administrator, instructed DNRC personnel to assume the responsibility of identifying potentially affected surface water sources for submitted applications. This meant that applicants were no longer expected to conduct that analysis themselves, but instead would be done by the agency as part of the Depletion Report. Given this change, Levens conceded the language of the Memo was not correct, but that, despite this intended change in the process, DNRC did not “have the capacity to do field investigations that would be needed” for the agency to identify all potentially connected surface waters. He offered the inconsistent explanation that some responsibility still lay with applicants to identify 11  potentially connected surface waters, stating “our analyses are based on the information we have available to us . . . it’s not our role to go out and do [the] work . . . ultimately, it’s still the applicant’s burden of proof to identify the surface waters, but . . . we have taken on that role . . . to do what we can do with the information we have to identify surface waters.” Olsen testified that, despite the Memo’s language, DNRC “still has the duty to assess whether any sources would be potentially impacted.” Importantly, Olsen’s understanding of the DNRC’s responsibility correctly reflects Admin. R. M. 36.12.1704(2): The department will identify the existing legal demands on the source of supply and those waters to which it is tributary and which the department determines may be affected by the proposed appropriation . . . (a) For groundwater appropriations, this shall include identification of existing legal demands for any surface water source that could be depleted as a result of the groundwater appropriation. (Emphasis added.) ¶14 However, no one at the agency ever attempted to identify or evaluate other potentially connected surface waters that could be impacted by Artesian’s pumping, nor did anyone seek this information from Artesian. Proceeding without such an inquiry, Folnagy completed the Aquifer Test Report and Depletion Report, which were reviewed by Levens, and then forwarded to Ward. ¶15 In turn, Ward utilized the Aquifer Test Report and the Depletion Report to conduct his analysis of the requisite criteria enumerated in § 85-2-311, MCA (-311 criteria). Those criteria include, among other factors, physical availability of the appropriated water, legal availability, lack of adverse impact to senior water right holders, and a beneficial use for 12  the water. Because physical availability and adverse impact were addressed in the Aquifer Test Report and the Depletion Report, Ward conducted only a legal availability analysis, wherein he concluded Artesian’s pumping would not impact Flathead River and Flathead Lake water levels. He did not conduct a legal availability analysis for any other surface water source. ¶16 On September 25, 2015, Artesian responded to Ward’s sole deficiency letter, providing the information requested. Having received that information, and with the physical and legal availability analyses completed, Ward concluded that Artesian had proven, by a preponderance of the evidence, that each of the -311 criteria were satisfied. Shortly thereafter, Ward authored, and Olsen reviewed and approved, a Preliminary Decision to Grant (PD) which the DNRC issued on January 14, 2016, and notified the public accordingly. The PD stated that, pursuant to the Memo, “[n]o additional modeling, evaluation of the zone of influence or aquifer flux calculations are needed . . .” Contested case hearing ¶17 Upon publication of the PD, Appellees Flathead Lakers and other groups and individuals (Objectors), including the United States Fish and Wildlife Service (USFWS), filed thirty-nine objections to Artesian’s application. In accordance with MAPA, a contested case hearing was held on September 19-21, 2017. The Hearing Examiner received pre-filed expert testimony from Artesian, USFWS, and the Objectors, for which Levens was designated by DNRC to review and respond. ¶18 Objectors’ pre-filed expert testimony included a report by hydrologist Dr. Shlomo Neuman, who had co-authored the Solution. Neuman opined that DNRC has “fail[ed] to 13  provide sufficient credible information” for its conclusions about Artesian’s proposed use. In Neuman’s opinion, the Solution had been improperly applied in analyzing the impacts of Artesian’s well because the site conditions were “fundamentally different than those underlying the [Neuman-Witherspoon] solution.” Specifically, Neuman offered that the Solution is intended for situations in which the well fully penetrates a uniform deep aquifer and draws water strictly horizontally.17 Neuman also opined that the aquifer test data that DNRC had inputted into the Solution was improper because the aquifer testing did not begin under hydrostatic conditions, but rather after extensive pumping from Artesian’s well had already caused significantly different initial water levels in the areas of the Koch and Nickol wells.18 ¶19 Hydrologist Dr. Willis Weight also submitted expert testimony for Objectors, and agreed with Neuman that the pumping of Artesian’s well for months prior to the aquifer test had undermined DNRC’s analysis. Weight calculated that the Nickol observation well is completed within the shallow aquifer, while the Koch observation well and Artesian’s well are completed within the top of the Deep Aquifer. Based on this alignment and the 17 Neuman’s solution assumes that the test well “fully penetrates” the aquifer, meaning that it is located towards the bottom of the aquifer. Here, Artesian’s well is located towards the top of the Deep Aquifer, and is not “uniform” because it is comprised of several different layers of varying types of soil which impact the flow of groundwater. 18 Generally, to measure the impact of groundwater pumping on local water levels, drawdown is calculated by measuring the change between the static background water level prior to pumping and the new, lower, stabilized water level during pumping. Because Artesian’s pump ran prior to the aquifer test, any measured background water level in the days prior to the test would not have reflected the static background water level prior Artesian’s well being pumped at all, but rather the stabilized water level resulting from Artesian’s initial 175 gpm pumping. 14  drawdown experienced at the observation wells during the aquifer test, Weight concluded that the top of the Deep Aquifer is connected to the shallow aquifer in the vicinity of Artesian’s well. Using his own solution to model the impact of Artesian’s pumping, Weight opined that Artesian’s pumping would reduce the quantity of water available for the surface spring water right of Objectors John and Amy Waller (Wallers) such that the Wallers would be unable to appropriate the full quantity of their water right. The Wallers own property across the road from Artesian’s facility, and their irrigation system is fed by a surface spring. ¶20 Dr. Tom Myers, expert for Objectors, opined that Artesian’s pumping “will cause drawdown in shallow aquifers, and affect surface water resources connected to that aquifer . . . [including] Egan slough.” Dr. Myers explained that the rate of drawdown and recovery levels measured in the Nickol well during Artesian’s tests “reflects a slow propagation of pumping stresses through layers . . .” and demonstrates that Artesian’s pumping “will draw water from shallow aquifers and will eventually cause a drawdown that expands through the shallow aquifer zone creating a capture zone of wells affected by [the pumping].”19 ¶21 USFWS’s expert, agency hydrologist Jaron Andrews, expressed concern that Artesian’s pumping would adversely affect several wells used by USFWS at the nearby Creston National Fish Hatchery. Andrews opined that the inadequacies of Artesian’s 19 Put another way, Dr. Myers opined that the delayed reaction of the Nickol well demonstrates the existence of a dividing aquitard between the Artesian and Nickol wells, thereby slowing the propagation of drawdown from one well to the other. However, the aquitard is not impervious, as demonstrated by the impact of Artesian’s well still had on the Nickol well. 15  aquifer test data could result in an underestimation of drawdown effects on local wells, and that Artesian’s pumping would likely reduce or eliminate the artesian flow currently present in the fishery’s several wells, resulting in a need for USFWS to purchase larger pumps. Another expert for Objectors, Mikel Siemens, likewise underscored the data omissions in Artesian’s aquifer test data, stated that the observation wells were not completed in the same water bearing zone or aquifer as Artesian’s well, and opined that Artesian’s well test was invalid because the well ran for eighty days prior to the test, making it impossible to accurately measure the static conditions of the aquifer prior to the test. ¶22 Brad Bennett and Roger Nobel20 submitted expert testimony for Artesian. Bennett opined that based on the Memo and the sizable water influx into the Deep Aquifer, Artesian’s pumping would not adversely impact prior appropriator’s ability to exercise their water rights. Noble opined that any potential adverse impacts of Artesian’s pumping are mitigated by Artesian’s ability to regulate and reduce their volume of water use during times of water shortages, should a senior user make a valid call. ¶23 In his report responding to the expert testimony, Levens acknowledged the Objectors’ concerns about issues with the aquifer testing data, i.e., the lack of background groundwater level data, the uncertainty surrounding the observation wells, and the initiation of pumping prior to beginning the well testing. Nevertheless, Levens found 20 Both Bennett and Noble were hydrologists employed with Applied Water Consulting, the company that Artesian contracted to complete its aquifer test. 16  Artesian’s aquifer test adequate. He offered that, notwithstanding the problem with the Nickol well’s use as an observation well, the observation well requirements were satisfied because the Koch observation well was completed in the same aquifer. In response to Dr. Neuman’s testimony, Levens acknowledged that Artesian’s aquifer test data “is not adequate to allow a comprehensive or rigorous analysis” using the Solution, but that DNRC had found the Neuman-Witherspoon solution to have the least shortcomings after considering other models. Levens offered that he was “unaware . . . of DNRC requiring the type of detailed investigation recommended by Dr. Neuman that would be necessary to rigorously apply the Neuman and Witherspoon (1969) model,” outside of controlled groundwater areas. Regarding Weight’s model, Levens stated that it had not been peer reviewed.21 ¶24 Folnagy, author of the Depletion Report, conceded the Application did not satisfy all regulatory requirements, but that he considered the information provided in Form 633 to be sufficient to conduct his analysis. Folnagy agreed his duties included identification of potentially affected surface waters,22 but stated the Memo “limits the scope of what [he] looked at for legal availability” because he applies it whenever a well is deemed to be within the Deep Aquifer. While offering that, “[i]f we determine there’s another potentially 21 Expert Dr. Thomas Maddock, III, testified that he agreed with Weight’s solution model. 22 In the methodology section of the Depletion Report, Folnagy explained the relevancy of net depletion to the agency’s legal availability analysis, and stated that “[n]et depletion is evaluated in three steps: identification of potentially affected surface waters, calculation of consumption, and calculation of the rate and timing of depletion to the identified affected surface waters.” (Emphasis added.) 17  affected source, we’ll look at it,” he was unclear about how he determines whether other water sources are potentially connected. Folnagy conceded there were inconsistencies between the conditions required for application of the Solution and the conditions present at Artesian’s well, namely, the well’s partial penetration of the aquifer, the heterogeneity of the aquifer, and the variance created by not accounting for vertical water draw, but believed use of the Solution here was still appropriate. ¶25 In his analysis that placed Artesian’s well in the Deep Aquifer, Folnagy estimated the top of Deep Aquifer to begin at 195 feet deep. On cross examination, he conceded that a 2004 study by the Montana Bureau of Mines and Geology depicted the Deep Aquifer as beginning at between 300 and 400 feet deep. Moreover, Folnagy estimated the thickness of the confining unit near Artesian’s well to be “between 200 to 300 feet thick,” but did not reconcile this measure of thickness with his estimation that the Deep Aquifer began at a depth of 195 feet. He also noted that in areas with a thin confining unit another surface water body could be the source of the Deep Aquifer’s recharge.23 Folnagy was clear that he did not identify any potentially connected surface waters other than the Flathead River and Flathead Lake, had not considered shallow aquifer wells in his legal availability analysis, did not review water right abstracts for surface waters, and did not model the effect of pre-stream capture. 23 Notably, 3D modeling included in Bennett’s pre-filed expert testimony indicates that the confining layer is 50 to 100 feet thick around Artesian’s well, significantly thinner than Folnagy’s estimated 200 to 300 feet thick. 18  ¶26 Folnagy and Levens both agreed that Egan Slough, nearby artesian springs, and other surface waters are within the projected cone of depression and therefore could be affected by Artesian’s pumping. Nevertheless, Ward confirmed that, for this application, DNRC “only looked at availability in the river and the lake” and that DRNC did not “compare the physical water supply for any surface water source in which water could be reduced by any amount as a result of [Artesian’s] well other than the Flathead River and Flathead Lake.” Folnagy confirmed that he did not consider impacts upon wells that were shallower than 100 feet. ¶27 Olsen confirmed that no variance was issued to Artesian despite the purpose of a variance being to allow applicants to circumvent information omissions akin to those occurring in Artesian’s application. Olsen also confirmed that the impact analysis on Deep Aquifer wells did not account for the potential reduction in artesian pressure in those wells, only the presence of water. Final Order and judicial review ¶28 Following the hearing, the Hearing Examiner (HE) issued the Final Order affirming DNRC’s decision to grant Artesian the water use permit. The HE found the Form 633 undisputedly omitted information, but that the missing information was “not material to the determination of the physical availability criteria,” reasoning that the incomplete Form 633 primarily impacted DNRC’s determination of whether the Application is correct and complete. The HE concluded that DNRC properly “considered the information provided by the Applicant as substantial credible information which included all the necessary information to evaluate the Application.” 19  ¶29 The HE characterized the Memo as limiting legal availability analysis of appropriations from the Deep Aquifer to only the Flathead River and Flathead Lake because only those water bodies “could be depleted [by groundwater pumping] absent specific information to the contrary.” Regarding legal availability, the HE found it was undisputable that “[g]round water is legally available” because the Deep Aquifer has an estimated 190,000 acre-feet per year of unappropriated groundwater influx, and concluded that “[d]epletions from surface water due to pumping of [Artesian’s] well at the proposed rate and volume of diversion are legally available to [Artesian].” That conclusion was supported only by a citation to the PD and the HE’s general factual findings. No mention is made of any effort to evaluate other potentially affected surface water sources by DNRC or Artesian. The HE concluded that “Objectors’ arguments regarding the completeness of the documentation of the pump/aquifer test and/or the appropriateness of the use of the Neuman-Witherspoon solution to develop aquifer properties, without providing any independent evidence that there is not groundwater or surface water legally available, are inadequate to overcome applicant’s proof by a preponderance of the evidence that both ground and surface water are legally available in the amount it seeks to appropriate.” ¶30 Appellees sought judicial review of the Final Order. After the District Court’s initial decision and this Court’s remand in Flathead Lakers I, the District Court found DNRC’s methodology lacking and reversed the Final Order on the grounds that the HE’s conclusions regarding the physical and legal availability of water, and the potential for adverse effects in water quantity and quality, were clearly erroneous. Regarding physical availability, the District Court was critical of DNRC’s use of the incomplete Form 633 as 20  the basis for the Aquifer Test Report and Depletion Report, stating that the omissions violated MWUA and “may have created an inaccurate picture of water availability and may also have caused inaccuracies throughout the entire -311 criteria review process.” The District Court also took issue with the HE’s categorization of the missing Form 633 data as “not material” to DNRC’s physical availability analysis, and instead concluded that the missing information caused DNRC to miss “a critical foundational step in determining whether a permit should be granted.” As such, the District Court concluded that the HE “arbitrarily found [Artesian’s] application established physical availability . . .” in a decision “unmotivated based on the record.” ¶31 Regarding legal availability, the District Court relied on the hearing testimony of Folnagy and Levens to conclude that DNRC’s reliance on the Memo, and the analysis of only the Flathead River and Flathead Lake, meant DNRC had not fulfilled its statutory obligations under MWUA and Admin. R. M. 36.12.1704-1706. The District Court concluded that although the Memo serves as “departmental guidance for reviewing whether a permit application in the deep aquifer will potentially affect surface water sources . . . in this case DNRC treated the memo as conclusory evidence that only the Flathead Lake and River could be affected” by Artesian’s pumping. The District Court noted the changing of the burden for identifying potentially affected surface waters over time, and reasoned DNRC erroneously operated on the procedure that if the applicant did not supply information about other potentially affected surface waters, then they did not need to conduct any further analysis. The District Court concluded that DNRC, by neglecting to consider surface water sources aside from the Flathead River and Flathead Lake, had failed 21  to properly assess the existing legal demands on water sources that “may be affected” or “could be depleted by any amount” due to Artesian’s proposed appropriation. The District Court vacated the Final Order. The District Court’s post-judgment Fee Order is discussed later herein. ¶32 Artesian and DNRC appeal. STANDARD OF REVIEW ¶33 This court applies the same standard of review as the lower court. DeBuff v. Mont. Dep’t of Nat. Res. & Conservation, 2021 MT 68, ¶ 22, 403 Mont. 403, 482 P.3d 1183. Judicial review of final agency decisions is governed by the Montana Administrative Procedure Act (MAPA). Upon review, a court “may not substitute its judgment for that of the agency as to the weight of the evidence on questions of fact.” Section 2-4-704(2), MCA. It may, however, “reverse or modify the decision if substantial rights of the appellant have been prejudiced because: (a) the administrative findings, inferences, conclusions, or decisions are: (i) in violation of constitutional or statutory provisions; (ii) in excess of the statutory authority of the agency; (iii) made upon unlawful procedure; (iv) affected by other error of law; (v) clearly erroneous in view of the reliable, probative, and substantial evidence on the whole record; (vi) arbitrary or capricious or characterized by abuse of discretion or clearly unwarranted exercise of discretion.” Section 2-4-704(2)(a). ¶34 An agency decision is clearly erroneous if, upon review of the record, it either: (1) is unsupported by substantial evidence upon review of the record, (2) misapprehends the effect of the substantial supporting evidence, or (3) “leaves the court with the definite and 22  firm conviction that a mistake has been made.” DeBuff, ¶ 23; Weitz v. Montana Dep’t of Natural Resources & Conservation, 284 Mont. 130, 133-34, 943 P.2d 990, 992 (1997). An agency’s decision is arbitrary or capricious if the decision is “at odds with the information gathered” or the product of internally inconsistent analysis. Mont. Envtl. Info. Ctr. v. Mont. Dep’t of Envtl. Quality, 2019 MT 213, ¶ 26, 397 Mont. 161, 451 P.3d 493. ¶35 A district court reviews an agency’s conclusions of law for correctness; this Court, in turn, applies the same standard on appeal. Flathead Lakers I, ¶ 7. This Court must review the entire administrative record when considering petitions for judicial review. DeBuff, ¶ 24. During review of administrative rulings, “[t]he Court’s focus is on the administrative decision-making process rather than the decision itself.” Park Cty. Envtl. Counsil v. Mont. Dep’t of Envtl. Quality, 2020 MT 303, ¶ 18, 402 Mont. 168, 477 P.3d 288. On highly technical matters and those requiring scientific expertise, we grant great deference to agency expertise. DeBuff, ¶ 24. To that end, “[t]his Court acknowledges that it is not comprised of hydrologists, geologists, or engineers, and that protecting the quality of Montana’s water requires significant technical and scientific expertise beyond the grasp of the Court.” Mont. Envtl. Info. Ctr., ¶ 20. We do, however, still retain the inherent power to review administrative proceedings to ensure that “agency decision-making is scientifically-driven and well-reasoned” and we require the agency be able to “cogently explain why it has exercised its discretion in a given manner.” Mont. Envtl. Info. Ctr., ¶¶ 20, 97 (quoting Nat’l Parks Conservation Ass’n v. United States EPA, 788 F.3d 1134, 1142 (9th Cir. 2015)). As such, “we will consider whether an agency decision was based on a consideration of all relevant factors.” Clark Fork Coal. v. Mont. Dep’t of Envtl. 23  Quality, 2008 MT 407, ¶ 21, 347 Mont. 197, 197 P.3d 482 (quoting North Fork Pres. Ass’n v. Dep’t of State Lands, 238 Mont. 451, 465, 778 P.2d 862, 871 (1989)). DISCUSSION ¶36 1. Did the District Court err by determining DNRC erroneously granted Artesian’s application for a beneficial use permit? ¶37 The Montana Water Use Act (MWUA) tasks the DNRC with “coordinat[ing] the development and use of the water resources of the state so as to effect full utilization, conservation, and protection of its water resources.” Section 85-1-101(3), MCA; Flathead Lakers I, ¶ 8. DNRC is therefore the exclusive provider of post-1973 water use permits, and processes applications for new appropriations. Section 85-2-302(1), MCA; Flathead Lakers I, ¶ 8. Within that role, DNRC must determine whether an application is “correct and complete.” Sections 85-2-102(9), -302(4)(a), MCA. Should the DNRC find an application to be deficient, “[t]he department shall notify the applicant of any defects in [the] application within 180 days.” Section 85-2-302(5), MCA. “If the department does not notify the applicant of any defects within 180 days, the application must be treated as a correct and complete application.” Section 85-2-302(5), MCA. This was crux of our decision in Flathead Lakers I—that Artesian’s application was deemed to be “correct and complete” by operation of § 85-2-302(5), MCA. ¶38 Once an application has been deemed to be “correct and complete,” DNRC must determine whether the applicant has satisfied the statutory criteria for issuance of a permit, under § 85-2-311, MCA. The burden lies with the applicant to prove, by a preponderance 24  of the evidence, that each -311 criterion is satisfied. Section 85-2-311(1)(a)(ii), MCA. Failure to establish any of the criteria necessitates denial of the application. ¶39 At issue in this appeal are the -311 criteria of physical availability and legal availability.24 Water is physically available if it exists “at the proposed point of diversion in the amount that the applicant seeks to appropriate.” Section 85-2-311(1)(a)(i), MCA. Water is considered legally available if the amount of physically available water exceeds the existing legal claims to that water. Section 85-2-311(1)(a)(ii), MCA. As further explained in regulation: 36.12.1705 PERMIT APPLICATION CRITERIA - COMPARISON OF PHYSICAL WATER AVAILABILITY AND EXISTING LEGAL DEMANDS (1) To determine if water is legally available, the department will compare the physical water supply at the proposed point of diversion and the legal demands within the area of potential impact. (2) For groundwater appropriations, in addition to (1) the department will compare the physical water supply for any surface water source in which water flow could be reduced by any amount as a result of the groundwater appropriation and the legal demands within the area of potential impact. Admin. R. M. 36.12.1705 (emphasis added). Pursuant to the statute and regulation, legal availability of the proposed use is determined by subtracting existing legal demands from physically available water at the proposed diversion site, or in short form: 24 Legal The District Court addressed these criteria, but also concluded that DNRC’s analysis of adverse impact, another criterion, was improper based upon a prior district court ruling regarding Artesian’s discharge permit issued by the Department of Environmental Quality. Our decision in Flathead Lakers v. DEQ, 22-0212, released concurrently with this Opinion, overturns the prior ruling upon which the District Court relied herein. Further, the analysis of adverse impact could be affected by the issues addressed herein. For that reason, we do not consider adverse impact in this Opinion. 25  Availability = Physical Availability – Existing Legal Demands.25 If there is not sufficient water for the proposed use after subtracting existing legal demands from the physically available water, then legal availability is not established. Section 85-2-311, MCA. A. Physical Availability ¶40 DNRC and Artesian argue the District Court contravened this Court’s directive from Flathead Lakers I and erred by concluding that the Form 633 omissions, see Admin. R. M. 36.12.121 (“minimum information that must be submitted with application”), invalidated DNRC’s physical availability analysis. Appellants argue the HE’s findings should be confirmed because they are supported by substantial evidence and are not error under the deferential clearly erroneous standard. The HE credited Folnagy’s testimony that, despite the Form 633 omissions, the omitted data did not hinder his ability to conduct the proper analysis, and DNRC was able to adequately assess the impact of Artesian’s pumping. Based on Folnagy’s testimony, the HE found the omitted Form 633 data is “not material to the determination of the physical availability criteria.” The HE reasoned that Objectors had conflated the missing data with satisfaction of the -311 criteria and that Admin. R. M. 36.12.1703, cited by Objectors, implemented the “correct and complete” process, not the -311 criteria. The District Court held the HE’s finding that the data omissions were not material to be clearly erroneous, and Appellants contend the District Court erred for several reasons, including, “again us[ing] the lack of compliance with Form 633 . . . was not the directive from this Court on remand.” 25 This is simplified for ease of understanding, but is not a true algebraic formula. 26  ¶41 However, Appellants clearly over-argue our holding in Flathead Lakers I. We did not hold the statutory “correct and complete” declaration had rendered the Form 633 omissions a closed chapter in the permitting process. Indeed, we explained the opposite: “The issue remains whether DNRC’s consideration of the application without the additional aquifer testing was arbitrary and capricious and whether its evaluation was clearly erroneous in light of the record.” Flathead Lakers I, ¶ 20 (emphasis added). Likewise, the HE’s determination that the Form 633 omissions only implemented the “correct and complete” process, and were not relevant to a determination of the -311 criteria, shortchanged the regulatory purpose of that data. The aquifer data is first required, under Admin. R. M. 36.12.121, to be submitted with the application. Admin. R. M. 36.12.1601(2) provides that “[a]n application deemed correct and complete can advance to the next stage of the application process.” Then, under Admin. R. M. 36.12.1703, entitled “Permit Application Criteria – Physical Ground Water Availability,” the agency is charged with conducting an “evaluation of drawdown in the applicant’s production well for the maximum pumping rate and total volume requested in the permit application using the information provided from the aquifer test.” Admin. R. M. 36.12.1703(2) (emphasis added). DNRC must “compare the drawdown projected for the proposed period of diversion to the height of the water column above the pump in the proposed production well to determine if the requested appropriation can be sustained.” Admin. R. M. 36.12.1703(3). If so, then water is considered physically available within the proposed well. Clearly, the data provided, and the data’s sufficiency, remains a significant consideration throughout the permitting process, including the assessment of the -311 27  criteria, and we concur with the District Court that the HE’s ruling that the data omissions were material only to the application’s correctness and completeness was an incorrect statement of law. ¶42 Artesian’s application was plainly noncompliant with Admin. R. M. 36.12.121. Fields on Form 633 were left blank and gaps for as long as thirteen hours exist in the recording of discharge rates during the well test. Levens testified that the minimum information standards “are designed to give [DNRC] the best shot [it] can to have a good aquifer test,” and stated that noncompliance with the minimums could result in applicants having a “bad aquifer test” resulting in subsequent DNRC analysis based on those tests being less reliable. Folnagy admitted he completed the Aquifer Test Report and Depletion Report based on partial information, “even though it didn’t comply with the law,” and even though a variance from the Application’s requirements had not been granted, Admin. R. M. 36.12.1703(4), based upon his determination the information provided was sufficient. ¶43 There were other issues beyond the incomplete Form 633. Accurate assessment of a groundwater appropriation’s impact requires accurate measurement of the area’s background water levels, which is properly done under hydrostatic conditions. Artesian’s well was allowed to pump for nearly three months prior to commencement of the aquifer test. Folnagy completed the Depletion Report under the mistaken understanding the well had been pumping for only about a month prior to the aquifer test. Several of Opponents’ experts opined that this discrepancy compromised DNRC’s measurements of background water levels because testing had not been performed upon the usual hydrostatic conditions. 28  ¶44 Objectors’ experts also asserted that the Nickol well, one of two planned observation wells used during the aquifer test, likely did not exist within the same water bearing zone as Artesian’s well, given the Nickol well’s apparent shallower depth and its unexpected, delayed reactions to the Artesian’s pumping. As noted above, the indication from the record was that the Nickol well was likely about 90 feet in depth and, based upon the testing response, was outside the Deep Aquifer, a possibility that DNRC conceded. It was thus disqualified as an observation well. The agency and the HE relied upon the singular Koch observation well, allowed under Admin. R. M. 36.12.121(3)(h), but further difficulties remained. ¶45 Dr. Neuman, co-author of the 1969 Solution used by Folnagy to model the impact of Artesian’s pumping, testified that Folnagy’s use of the Solution was improper. According to Neuman, the conditions in the aquifers at the site of Artesian’s well differed significantly from the conditions required for application of the Solution, and that those differences were significant enough to invalidate Folnagy’s analysis. Instead, Neuman stated that DNRC should have conducted thorough groundwater modeling, including three-dimensional modeling among other approaches, to determine the actual characteristics of the underlying aquifers. DNRC conceded that the Solution was not a perfect fit but reasoned that it was their best available option, and opposed Neuman’s recommendation for analysis because it “would be very costly and was not required historically of any other applicant regarding aquifer tests.” Levens also noted that such extensive groundwater analysis is not typically required “except where water availability or adverse effect concerns have been expressed in a petition for a controlled groundwater 29  area.” Even assuming the potentially flawed testing was sufficient, Neuman also opined that the aquifer test data inputted into the Solution was improper because the aquifer testing did not begin under hydrostatic conditions, but rather after extensive pumping from Artesian’s well had already caused significantly different initial water levels in the areas of the Koch and Nickol wells. ¶46 In summary, to test the impacts of Artesian’s well, DNRC applied the Solution, despite the Solution’s tenuous application to the conditions at the site, which, in differing degrees, all parties acknowledged. Into the Solution DNRC inputted data that was partially incomplete (Form 633 omissions), potentially suspect (extensive pre-test pumping altering static conditions), and based on an incorrect assumption (that the well had only been pre-test pumped for less than a month). Artesian utilized a flow totalizer, from which it deduced, and which DNRC and the HE accepted as an alternative measure, an average flow rate over the course of the testing. While acknowledging Objectors had raised legitimate concerns about the lack of background groundwater level data, uncertainty about the observation wells, and the pre-test pumping, Levens testified he believed the aquifer testing was nonetheless adequate. Artesian’s experts, Bennett and Nobel, opined that, given the size of the Deep Aquifer, Artesian’s pumping would not adversely impact prior appropriator’s ability to exercise their water rights. ¶47 If this were the end of the matter, that is, if the only issue was determination of physical availability, we could conceivably conclude, applying appropriate deference to the agency’s expertise, that under the first prong of the clearly erroneous standard of review there was substantial evidence to support the HE’s findings on physical availability, despite 30  significant contrary evidence. That would take us to the second and third prongs of the standard, whether, despite the substantial evidence, the HE “misapprehended the effect of the evidence,” or whether “a review of the record leaves the court with a definite and firm conviction that a mistake has been made.” DeBuff, ¶ 23. However, we need not make that determination at this juncture, because, as explained above, in this case physical availability is not an end to itself, but is one factor within a formula to determine legal availability, and thus has further significance to the ultimate conclusion, discussed below. B. Existing Legal Demands and Legal Availability ¶48 Appellants argue the District Court erred by holding DNRC improperly “treated the memo as conclusive evidence that only the Flathead Lake and River could be affected by [Artesian’s] operation,” and by concluding the HE’s findings based on the Memo were clearly erroneous, including that the only surface water sources “which could be depleted absent specific information to the contrary” were the Flathead River and Flathead Lake, for which any depletion was legally available to Artesian. DNRC argues its decision to use the Memo and only analyze surface legal availability for the Flathead River and Flathead Lake was properly based on the lack of any evidence indicating other potentially affected water sources, and stresses the agency’s expertise, stating that “specialized knowledge comes in different forms, including for this case a memorandum on the characteristics of the Deep Aquifer.” Objectors argue the District Court correctly determined DNRC had failed its duty to fully analyze the legal availability of the water Artesian sought to appropriate because DNRC only evaluated the Flathead River and Flathead Lake. 31  ¶49 Regarding legal availability, the -311 criteria provide that: (ii) water can reasonably be considered legally available during the period in which the applicant seeks to appropriate, in the amount requested, based on the records of the department and other evidence provided to the department. Legal availability is determined using an analysis involving the following factors: (A) identification of physical water availability; (B) identification of existing legal demands of water rights on the source of supply throughout the area of potential impact by the proposed use; and (C) analysis of the evidence on physical water availability and the existing legal demands of water rights, including but not limited to a comparison of the physical water supply at the proposed point of diversion with the existing legal demands of water rights on the supply of water. Section 85-2-311(1)(a)(ii), MCA (emphasis added). The regulations amplify this process, including the agency’s duty: 36.12.1704. PERMIT APPLICATION-EXISTING LEGAL DEMANDS […] (2) The department will identify the existing legal demands on the source of supply and those waters to which it is tributary and which the department determines may be affected by the proposed appropriation. (a) For groundwater appropriations, this shall include identification of existing legal demands for any surface water source that could be depleted as a result of the groundwater appropriation. Admin. R. M. 36.12.1704 (emphasis added). ¶50 As noted above in the background discussion, the Memo advised the agency that applications for water use from the Deep Aquifer required a surface water legal availability analysis to be conducted only for Flathead River and Flathead Lake because “groundwater levels in the deep aquifer are effectively controlled by the surface water levels in the Flathead River and Flathead Lake.” The Memo cautioned, however, that there could be connections between the Deep Aquifer and “other surface waters or reaches of the Flathead 32  River,” and in such cases, “applicants need to evaluate availability on those sources.” While there was at times inconsistent testimony from DNRC employees, it is clear that, despite an internal agency effort to return the burden of identifying existing demands on other surface waters to the agency in 2012, and despite placement of the burden on the agency by Admin. R. M. 36.12.1704, quoted above, DNRC did not undertake that task here, and instead continued to rely on the Memo’s placement of the burden upon the applicant. ¶51 When drafting the Aquifer Test Report and Depletion Report, Folnagy applied the Memo as written, given Artesian’s well’s location in the Deep Aquifer, and identified only the Flathead River and Flathead Lake as “Potentially Affected Surface Waters.” Because, in accordance with the Memo, no other information had been provided by the applicant indicating other surface waters could be potentially affected, Folnagy listed no other “Potentially Affected Surface Waters” in the Depletion Report. ¶52 Ward used the Aquifer Test Report and Depletion Report to prepare his technical report, including the legal availability analysis on those waters identified in the Depletion Report as being potentially affected, the Flathead River and Flathead Lake. A legal availability analysis for other potentially affected surface waters was not conducted. Olsen reviewed Ward’s technical report, which was used as the basis for the agency’s Preliminary Decision to grant the permit. Olsen did not review Ward’s process for analyzing legal availability. Consequently, regardless of whom is to blame, the bottom line is that the agency, citing the Memo, deferred to the applicant and did not attempt to identify potentially affected surface waters aside from the Flathead River and Flathead Lake, a 33  violation of Admin. R. M. 36.12.1704(2)(a), which imposes a clear duty upon the agency: “The department will identify the existing legal demands on the source of supply . . . (a) For groundwater appropriations, this shall include identification of existing legal demands for any surface water source that could be depleted as a result of the groundwater appropriation.” (Emphasis added.) While § 85-2-311(1)(a)(ii), MCA, assigns the burden of proof to an applicant to prove that water “can reasonably be considered legally available” at the place of appropriation, it is the agency’s duty to determine what needs to be proven, i.e., what legal demands exist for which the applicant must prove reasonable legal availability. This is only logical: the applicant, here Artesian, has no incentive to identify additional potentially affected water sources for which it must prove legal availability. We cannot defer to an agency’s expertise when it has failed to satisfy its legal obligation. ¶53 Dr. Weight’s testimony included his own solution model which demonstrated that some portion of the nearby Waller surface spring water right would be captured by Artesian’s pumping. While DNRC rebutted Weight’s model as lacking peer review, the point is not whether Weight’s analysis of the impact on the Waller right was right or wrong; the point is that DNRC did not attempt to identify the Waller right as a potentially affected source nor consider it as an existing legal demand, in violation of its duty. ¶54 Nor did DNRC consider shallow wells in its analysis of potentially affected water rights. According to Folnagy, only wells at greater than 100 feet in depth were considered as potentially impacted and, consequently, he did not include shallow wells in his analysis. However, Admin. R. M. 36.12.175 requires DNRC to “compare the physical water supply at the proposed point of diversion and the legal demands within the area of potential 34  impact.” Based on the record, shallow wells appear to be within the area of potential impact. Folnagy testified that the Deep Aquifer is leaky, meaning that there is an interchange of water with the shallow aquifer, and he conceded that Artesian’s year-round pumping could draw water out of the shallow aquifer. During Artesian’s aquifer test, the Nickol well experienced drawdown despite its much shallower depth. Opponents’ experts testified that the Nickol well was probably in a shallow aquifer, given its shallow depth and its delayed response to Artesian’s pumping. Folnagy concluded no area wells of greater than 100 feet of depth would experience drawdown below their typical water column levels, yet this failed to consider wells under 100 feet in depth that could lie in a shallow aquifer, such as Nickol’s. ¶55 In conclusion, errors of law were committed during the agency’s processing of the application, including the failure to submit all the required data, and a failure to fulfill the agency’s duty to identify and analyze all potentially affected sources. The data omissions, particularly the background water levels and the well perforation levels, serve to erode the agency’s analysis of impacts, and the long-term pre-test pumping disturbed the well’s hydrostatic condition—the status quo—that was to exist at the start of testing. The data was inputted into a solution model that was tenuously applicable. These concerns in the determination of physical availability were then combined with DNRC’s failure to undertake its responsibility to identify other potentially affected surface water sources, in default to the Memo, which was a violation of its duty to ensure that all existing water demands are considered. 35  ¶56 The errors of law and process undermine confidence in the agency’s determinations. Legal availability is founded upon the assessment of physical availability, which suffered from incorrect or omitted data, and existing legal demands, which suffered a failure of the agency to fulfil its duty to investigate further, despite the existence of an active surface right across the road. Consequently, this combination of deficiencies leaves us with the definite and firm conviction that, upon review of the whole record, and despite substantial evidence in favor, a mistake was made in the HE’s finding of legal availability. We thus concur with the District Court’s determination that the finding was clearly erroneous. ¶57 A significant impression drawn from the record is this: that because there is so much water in the Deep Aquifer, the agency assumed the proposed well would have little impact, and passed it along without diligent review. The agency may be right in the end, but until the proper assessment is done, the Objectors were all prejudiced by the agency’s failure to complete it. We conclude the District Court correctly vacated the Final Order. ¶58 2. Did the District Court err by denying Objectors’ motion for attorney fees? ¶59 The Objectors’ cross appeal of the deemed denial of their motion for attorney fees has reached the Court via a convoluted procedural pathway. On September 30, 2021, the District Court issued its Order vacating the HE’s Final Order and remanding the case to the DNRC. Objectors timely filed a motion for attorney fees, along with a motion to bifurcate the fee proceeding into entitlement and reasonable calculation components, on October 15, 2021. Artesian and DNRC timely submitted their individual briefing in opposition to the motion for fees. On October 29, 2011, Artesian filed a Notice of Appeal of the merits order. On November 1, 2021, the District Court issued an Order granting 36  Objectors’ motion to bifurcate the fee matter to “address the issues of entitlement to fees separately from the reasonable amount to be awarded.” ¶60 On January 3, 2022, Objectors moved this Court to dismiss Artesian’s appeal, arguing it was unripe because the District Court had not yet entered a fee order. DNRC opposed the motion, and we denied it on January 25, 2022, reasoning that Artesian’s appeal was ripe because, as of December 13, 2021, under then-applicable rules, Objectors’ motion for fees was “deemed denied” as a matter of law after expiration of the 60-day window for ruling under Rule 59(f). Citing Ballou v. Walker, 2017 MT 197, ¶ 22, 388 Mont. 283, 400 P.3d 234, we explained, “[t]his Court treats a motion for attorneys’ fees that is filed after a judgment is entered as a motion to alter or amend the judgment,” and, as such, Rule 59(f) applies, and any such motion not ruled upon within 60 days of its filing date is deemed denied. See Ballou, ¶ 22 (citing Associated Press v. Crofts, 2004 MT 120, ¶¶ 36-37, 321 Mont. 193, 89 P.3d 971).26 Despite our order denying the motion to dismiss, the District Court, on February 14, 2022, issued its “Order on the Motion for Attorney Fees” (Fee Order). Therein, the District Court held Objectors were entitled to attorney fees, and Artesian was obligated to pay fees, but that the Department was exempted from paying fees under application of quasi-judicial immunity. ¶61 Apparently recognizing the probable invalidity of the Fee Order granting them fees, Objectors filed, on February 24, 2022, a Notice of Cross-Appeal, not from the Fee Order, 26 This point was reiterated in our March 15, 2022, Order denying Objectors’ motion for rehearing of our denial of their motion to dismiss, and became the law of the case. 37  but from the “deemed denial” of their fee motion. However, their cross appeal from the deemed denial, which occurred by operation of law on December 13, 2021, was untimely, and DNRC moved to dismiss it. In a June 7, 2022, Order, this Court denied DNRC’s motion to dismiss and granted Objectors’ request for an out-of-time cross appeal. In that Order, we explained the case’s procedural history and concluded, given the “procedurally confusing situation in which neither the parties nor the District Court had a clear grasp on the appellate timeline,” that “extraordinary circumstances exist such that denying Flathead Lakers’ cross-appeal as untimely would be unjust.” Further, we referenced the subsequent revision of Rule 4, M. R. App. P., effective January 1, 2022, which would, going forward, avert the timing conundrum that occurred here. The new Rule provides: A notice of appeal filed prior to the district court’s ruling on any necessary determination of the amount of costs and attorney fees awarded, or sanctions imposed, may be dismissed sua sponte and shall be dismissed upon the motion of any party. The district court is not deprived of jurisdiction to enter its order on a timely motion for attorney fees, costs, or sanctions by the premature filing of a notice of appeal, in accordance with Rule 58(e), M. R. Civ. P. Rule 4(5)(a)(iii), M. R. App. P. (emphasis added). Under this revision, the District Court would not have lost jurisdiction for issuance of its Fee Order, and this Court would have had the option of dismissing Artesian’s appeal to permit issuance of the Fee Order. Thus, for these reasons, we granted Objectors an out-of-time cross appeal of the deemed denial of their fee motion.27 27 This rule revision also eliminates the distinction we recognized in Bitterroot River Protective Association v. Bitterroot Conservation District, 2011 MT 51, ¶¶ 13-14, 359 Mont. 393, 251 P.3d 131, of those cases wherein a district court entered a judgment that explicitly reserved jurisdiction to thereafter determine fees and costs. 38  ¶62 The procedural confusion did not end there, but was furthered by the parties’ appellate briefing. While acknowledging the Fee Order was void, Objectors’ opening cross-appeal brief nonetheless lauded the rationale of the Fee Order, arguing that, “[w]hile the District Court’s reasoning was correct, the timing of the decision was an abuse of discretion.” Thus, Objectors simply requested this Court to remand the case to provide additional time to act regarding fees, “consistent with the District Court’s February 14, 2022 Order.” Notably, Objectors’ brief made no mention that the Fee Order had, inter alia, exempted the Department from paying fees by application of quasi-judicial immunity, and offered no argument regarding any asserted obligation of the Department to pay fees. ¶63 Given Objectors’ arguments, DNRC was not put on notice by the opening cross-appeal briefing that fees were being sought, or would be sought, from the Department. Consequently, DNRC made no argument in its answer brief about an agency’s liability for fees, rather, argued the deemed denial of the fee motion occurred as a matter of law, that whether the District Court believed it had jurisdiction to rule on the motion was of no import, and that remand was improper. Citing Ballou, the Department argued that the Fee Order was void, and that the deemed denial should be affirmed. ¶64 In its answer to the cross appeal, Artesian resisted fees by arguing that an applicant is “shackled to decisions made solely by the Department” and “should not have to suffer for the Department’s incorrect policy and procedural decisions.” Then, Artesian argued that the fee statute, § 85-1-125, MCA, provided no exemption for the Department, which should be held to pay fees because it stepped beyond its quasi-judicial adjudication role, affirmatively undertook litigation as a party to defend the agency’s decision, took partisan 39  positions, and created costs for the other parties. Objectors, apparently taking a cue from Artesian, argued for the first time in their reply brief that the Department should be liable for fees “during the period of time DNRC acted as an adversarial party litigant and not a neutral decision maker.” ¶65 In response to these filings, and with no further opportunity under the Rules to respond, the Department filed a motion to strike the portions of Artesian’s answer brief and Objectors’ reply brief that argued the Department was responsible for fees. The Department noted that, not only did Objectors not raise the issue of the Department’s obligation to pay fees in their opening brief, they instead had signaled just the opposite by affirmatively arguing that the reasoning of the Fee Order—which had exempted the Department from fees—was correct (“[w]hile the District Court’s reasoning was correct, the timing of the decision was an abuse of discretion.”). The Department further noted that Artesian had not raised the issue of the Department’s obligation to pay fees in its appeal, raising the issue only in response to the cross appeal. Thus, the Department argued the Objectors and Artesian were “attempt[ing] to improperly raise this novel issue for the first time in cross-appeal briefing to which Appellant DNRC has no opportunity to respond,” and asked for these arguments to be stricken. We took the motion under advisement. ¶66 To begin, we clarify the issue to be resolved. This is an appeal from the District Court’s denial of Objectors’ fee motion, which occurred by operation of law. As we ruled in our previous orders in this case, and in previous cases, because the fee motion was deemed denied on December 13, 2021, the subsequently-entered Fee Order is void. Ballou, 40  ¶ 23. Objectors’ arguments rely on the rationale expressed in the Fee Order but, legally, that order does not exist, and it has no effect on this Court’s review of the denial of fees. ¶67 It is also important to clarify that this case is postured differently than Ballou and Crofts. In those cases, the parties against whom fees were awarded challenged the untimely, post-judgment orders by which the assessments had been made. Ballou, ¶ 21; Crofts, ¶ 35. We held that the fee orders in those cases, entered after the fee motions were deemed denied by operation of law, were void, and we reversed the fee assessments. Ballou, ¶¶ 22-24; Crofts, ¶¶ 34-38. Unlike those cases, here all parties acknowledge that the Fee Order is void, and no one is challenging its validity on appeal. Rather, it is the deemed denial of the fee motion that is challenged by Objectors, the party who was denied fees. Thus, the Department’s argument that fees must be denied on the basis of Ballou and Crofts is inapposite. ¶68 Objectors argue that “the gravamen of the issue presented here . . . is whether it was an abuse of discretion for the District Court not to have timely ruled pursuant to the motion brought under MCA § 85-2-125.” (Emphasis added.) However, this is an incorrect statement of the issue. To hold that a deemed denial by operation of rule occasioned by a district court’s failure to act is itself an abuse of discretion would negate Rule 59(f), because every deemed denial would necessarily constitute error. A deemed denial can occur for many reasons, including that a district court may decide to “pocket” deny the motion without expending the effort to write an order; or because the district court simply forgot about the motion. Regardless, whatever the reason for the lapse of the requisite time, the motion is denied as a matter of law. On appeal of that ruling, the issue is not whether the 41  district court erred by allowing the time to lapse and the deemed denial to occur. Rather, the issue is whether the district court’s deemed denial of the motion was error on its merits. See Detienne v. Sandrock, 2017 MT 181, ¶¶ 40-42, 388 Mont. 179, 400 P.3d 682; Green v. Gerber, 2013 MT 35, ¶¶ 30-43, 369 Mont. 20, 303 P.3d 729; Sun Mt. Sports, Inc. v. Gore, 2004 MT 56, ¶¶ 29-31, 320 Mont. 196, 85 P.3d 1286 (in each case this Court assessed whether, on the merits, a district court had abused its discretion by way of a deemed denial of a motion to set aside default or default judgment).28 ¶69 Section 85-2-125(1), MCA, governs this proceeding and provides, “[i]f a final decision of the department on an application for a permit . . . is appealed to district court, the district court may award the prevailing party reasonable costs and attorney fees.” (Emphasis added.) Prior to 2011, district courts possessed no discretion and were required to award fees, because the statute then provided that “the district court shall award the prevailing party reasonable costs and attorney fees.” Section 85-2-125(1), MCA (2009) (emphasis added). In 2011, the Legislature amended the statute by changing “shall award” to “may award,” thereby making discretionary a prevailing-party fee award in permit proceedings. Because the statute now makes the award permissive, subject to the discretion of the district court, we review a district court’s decision to grant or deny fees for abuse of 28 For the same reasons, Objectors’ argument based upon State v. Weaver, 276 Mont. 505, 917 P.2d 437 (1996), that “a court’s failure to exercise its discretion is, in itself, an abuse of discretion,” is inapplicable. Here, by operation of law, the court’s authority was indeed exercised, and a decision was made to deny the fee motion. 42  discretion. Baxter v. State, 2009 MT 449, ¶ 46, 354 Mont. 234, 224 P.3d 1211; Chase v. Bearpaw Ranch Ass’n, 2006 MT 67, ¶ 15, 331 Mont. 421, 133 P.3d 190. ¶70 Objectors argue that it is “beyond cavil” they prevailed in the litigation, and argue they were the prevailing parties because they obtained a “net benefit” in the Order on Remand by “achieving an outcome that included the District Court vacating the contested water use permit and remanding the matter back to DNRC,” and that now restricts Artesian “from continuing to appropriate groundwater under the permit at issue.” We note from a review of the proceeding that Objectors fully participated in the extensive hearing before the HE, initially prevailed in the District Court leading to the first appeal, and after this Court’s remand, prevailed a second time in the District Court. They have now prevailed in this appeal, after seven years of litigation. Objectors marshalled extensive expert testimony and addressed a voluminous record to support their claims. Objectors uncovered the errors in DNRC’s review process that led the District Court, and now this Court, to reject the permit, despite usual deference owed to the agency. After extensive effort, Objectors are clearly the prevailing party. We therefore conclude that Objectors are entitled to an award of fees and that the District Court’s denial of the fee motion, by operation of law, was an abuse of discretion. ¶71 That is the only fee issue properly before us—Objectors’ entitlement to fees under § 85-2-125, MCA, after the deemed denial of their motion. Other issues, such as Artesian’s and the Department’s respective obligations to pay fees, Artesian’s fairness-based arguments, and the amount that reasonably should be assessed, are more appropriately considered on remand, upon proper process. The Department’s argument in the motion to 43  strike is well taken—the issue of its responsibility for fees was not properly raised in the cross-appeal, and it had no opportunity to respond to the late-raised claim. See State v. Porter, 2018 MT 16, ¶ 16, n.1, 390 Mont. 174, 410 P.3d 955 (citations omitted) (“This Court will not entertain an argument first raised in a reply brief.”). As we do not address that issue here, we deny the motion as moot. ¶72 The District Court erred in denying, by operation of law, Objectors’ motion for attorney fees. We remand for consideration of remaining fee issues by the District Court. ¶73 Affirmed in part, reversed in part, and remanded for the further proceedings consistent herewith. /S/ JIM RICE We concur: /S/ MIKE McGRATH /S/ LAURIE McKINNON /S/ JAMES JEREMIAH SHEA /S/ BETH BAKER /S/ INGRID GUSTAFSON /S/ DIRK M. SANDEFUR 44