IEEE Power & Energy Magazine - September/October 2020 - 39

Hydrologic input data have to be carefully analyzed. They
were characterized using back-calculated inflow data derived
from 12 years of available historical records on generation,
spill flows, and water levels at the Acaray power plants. Given
the long-term nature of the analysis, consideration for climate change was also included. Forty-three years of natural
inflow projections at the various inflow points starting from
the current year of the study (2017) and extending to the year
2060 were used to complement the hydrologic representation
derived from historical records. The projections were calculated on a daily time step based on future precipitation and
temperature inputs provided by climate models under a separate study.
Load projection was another key input to the analysis.
The maximum demand is expected to grow significantly
over the next few years; the current load is predicted to double by 2025. To consider uncertainty in the forecasts over the
very long term, an extensive sensitivity analysis was carried
out to evaluate the impact of higher and lower demands. Seasonal variation in load patterns was also considered.

✔ annual spending limits
✔ allowable or blocked windows, constraining the year

in which an intervention can take place
✔ precedence relationships between interventions including
YGAQ
YGAW
Yguazú
Reservoir
YGAT

YGARR

CONQ
CONRR
ACYQ
ACY1
ACY2
ACYT

ACYS

ACYRR

Asset Management Model

The purpose of the asset management model (HydroVantage) was to calculate the operating costs of the existing
facility with and without future upgrade options. The model
was designed to combine costs with the generation revenues
derived from the Power Generation Planning model to arrive
at a cost-benefit comparison through the evaluation of the net
present value (NPV).
The model used in this analysis was specifically designed
to perform a risk-based capital expenditure (CAPEX) optimization for hydroelectric power plants. It uses statistically
derived failure probability curves, and its benchmark database
contains costs, outage durations, and rates for various typical
component failure and intervention modes. Its primary output
is the intervention schedule that provides the lowest overall
NPV of CAPEX, outage (risk) cost, and operation and maintenance (O&M) costs minus any performance benefits associated with an intervention. This is illustrated in the formulation
of the economic objective within the model where the NPV is
calculated by summing the terms noted previously (i.e., NPV
= risk + capital + O&M − benefits).
The main concept behind the asset management model
is presented in Figure 3, where upgrade options are first
evaluated using a risk-based analysis, followed by a prioritization scheme to consider other operating requirements.
The risk-based analysis is an unconstrained optimization that looks at only economic tradeoffs between the
various capital investments. The prioritization/levelizing
scheme considers various financial and operational constraints, thus redistributing various interventions from the
risk-based analysis over the given time period such that the
increase in an intervention schedule's NPV is minimized.
The prioritization/levelizing scheme can impose various
types of constraints:
september/october 2020

YGAZ
Legend
Inflow Arc

Source/Sink

Power Arc

Reservoir

Spillway Arc

River Junction
Watershed Element

River Reach Arc
(a)
System Load
Acaray
I

Itaipú, Capacity
Contract

Acaray
II

Itaipú Excess
Acaray
III
Yacyreta
Yguazú
(b)

figure 2. A hydropower and transmission system representation-the Power Generation Planning model. YGAQ:
Yguazú inflow; YGAW: Yguazú watershed; YGARR: Yguazú
to confluence; CONRR: Yguazú to Acaray; ACYS: Acaray spillway; YGAT: Yguazú outflow; CONQ: confluence
inflow; ACYQ: Acaray inflow; ACY1: Acaray I plant; ACY2:
Acaray II plant; ACYT: Acaray outflow; ACYRR: Acaray outlet; YGAZ: source/sink.
ieee power & energy magazine

IEEE Power & Energy Magazine - September/October 2020

Table of Contents for the Digital Edition of IEEE Power & Energy Magazine - September/October 2020