Scientific collaboration
Joint research, applied studies, academic cooperation, theses, publications and new LREM applications.
THEK Method · Reservoir Modeling · Open Science
LREMLOW-FLOW RESERVOIR EVOLUTION MODEL
A parsimonious empirical framework for analyzing reservoir behavior during low-flow periods through observable variables and explicit mathematical relationships. The method organizes the sequence FGET → VRR → NFF into the operational chain September elevation loss → Invernal Type → reservoir elevation Z(d) → potential rationing hours H.
Mazar ReservoirEcuadorHydropowerFGET · VRR · NFFOpen Science
DOI 10.5281/zenodo.23145585
01 · Scientific method
What LREM does
LREM describes the evolution of a regulating reservoir during a low-flow season and connects hydrological severity with an explicit reservoir-elevation trajectory and an empirical estimate of potential daily rationing hours. The development window uses real Mazar data from 2017–2025, while 2023 and 2024 provide direct verification cases and 2025 provides a stable reference condition.
ΔZsepSeptember elevation loss
ITInvernal Type severity factor
Z(d)Time-dependent reservoir elevation
HPotential rationing hours
FGET
Explores candidate relationships among observable variables: elevation–time, hours–time and hours–elevation representations.
VRR
Reduces and ranks variables, coefficients and scales so the dominant structure of the reservoir problem becomes explicit.
NFF
Reorganizes the reduced relations into a compact chain of linked equations that remains interpretable and reproducible.
02 · Mathematical architecture
Core formulation
The formulation keeps every stage visible. Inputs, transformations and outputs are explicit, making the method suitable for comparison, technical review and updating when new observations become available.
ΔZsep = Z1sep − Z30sepSeptember elevation loss.
IT = 1 + ΔZsep/16Dimensionless Invernal Type factor; higher September drawdown produces a larger severity value.
Z(d) = (−b0/2200) [ ITd²/141 − ITd + 6663.51 ], b0 = −705Time-dependent reservoir elevation. The parabolic minimum occurs at d = 70.5.
H = max(0, 0.05 IT3.7(2124 − Z))Empirical estimate of potential daily rationing hours; the factored form identifies 2124 m as the mathematical activation threshold.
03 · Verification and 2026 application
Observed agreement and operational interpretation
0.0029%relative elevation difference · 2023
0.0540%relative elevation difference · 2024
1.57%rationing-hours difference · 2023
0.12%rationing-hours difference · 2024
For the two hourly verification points, the mean absolute percentage difference is approximately 0.84%. In the 2026 application, the method separates the original LREM curve from a curve adjusted to actual 2026 observations. The original central scenario gives IT = 1.8831 and a minimum around 2114.08 m, while the adjusted curve yields a higher minimum of 2122.08 m and approximately 1.00 h/day at the adjusted minimum.
Scientific-use statement. LREM is a mathematical model and prospective reference, not a deterministic prediction of power outages. Its purpose is to model low-flow behavior so that scenarios, risk windows and possible operational needs can be analyzed transparently and updated with observations.
04 · Publications and source documents
Read the complete method
The complete LREM documentation is available here in English and Spanish. The Zenodo record is the open-science reference for the method and provides a persistent DOI for discovery, citation and reuse under its repository terms.
LREM Method · Mazar · EnglishFGET–VRR–NFF formulation, historical window 2017–2025, equations and verification.Método LREM · Mazar · EspañolFormulación FGET–VRR–NFF, ventana histórica 2017–2025, ecuaciones y comprobación.LREM · Mazar 2026 · V11 · English2026 application: original curve, adjusted curve, risk zones and operational reading.LREM · Mazar 2026 · V11 · EspañolAplicación 2026: curva original, curva ajustada, zonas de riesgo y lectura operativa.
06 · Collaboration, courses, advisory & licensing
Work with THEK Research Institute
THEK Research Institute invites universities, research centers, engineering faculties, public institutions, companies, professionals and scientific teams to collaborate around LREM and the broader THEK methodological ecosystem.
Courses & training
Specialized courses, seminars, master classes and technical training on LREM, FGET, VRR, NFF and related THEK methods.
Technical & academic advisory
Advisory services for reservoir modeling, low-flow analysis, risk scenarios, research design and methodological implementation.
Licensing
Institutional and professional licensing opportunities for authorized use, implementation and development of THEK methodologies.
Interested in collaborating with THEK?
For scientific cooperation, courses, advisory services, institutional projects or licensing, contact THEK Research Institute at research@hithek.com.
07 · Open science and machine discovery
Persistent scientific identity
Method: LREM — Low-Flow Reservoir Evolution Model
Creator: M.Sc. Marcelo Moncayo Theurer · THEK Research Institute
Zenodo record: zenodo.org/records/23145585
Canonical website: https://lrem.hithek.com/
Institutional contact: research@hithek.com
This page exposes standard scholarly metadata, DOI metadata, JSON-LD, hreflang, canonical links, robots directives, sitemap entries, AI-readable text files and direct PDF links so search engines, scholarly crawlers and AI systems can identify LREM as a named THEK scientific method and connect the website with its open Zenodo record.