controller ¶

Adjusts KLA values based on electricity prices and ammonia concentration in the effluent.

Controller ¶

Controller(price_percentile, klas_init, kla_reduction, s_nh_threshold, elec_price_path=None)

Creates a Controller object.

Parameters:

Name	Type	Description	Default
`price_percentile`	`float`	Percentile of electricity prices used to adjust KLa values (aeration reduction at prices above the percentile, e.g. 0.9 -> aerate less when electricity prices are in the top 10%).	required
`klas_init`	`ndarray`	Initial KLa values for the reactor compartments [d⁻¹].	required
`kla_reduction`	`float`	Reduction factor for KLa values.	required
`s_nh_threshold`	`float`	Maximum value of ammonia concentration in the effluent [g ⋅ m⁻³].	required

Source code in src/bsm2_python/controller.py

def __init__(
    self,
    price_percentile: float,
    klas_init: np.ndarray,
    kla_reduction: float,
    s_nh_threshold: float,
    elec_price_path: str | None = None,
):
    self.price_percentile = price_percentile
    self.klas_init = klas_init
    self.kla_reduction = kla_reduction
    self.s_nh_threshold = s_nh_threshold
    self.is_price_in_percentile: list[float] = []
    path_name = os.path.dirname(__file__)
    if elec_price_path is None:
        self.elec_price_path = path_name + '/data/electricity_prices_2023.csv'
    else:
        self.elec_price_path = elec_price_path
    with open(self.elec_price_path, encoding='utf-8-sig') as f:
        prices = []
        price_times = []
        data = np.array(list(csv.reader(f, delimiter=','))).astype(np.float64)
        for price in data:
            prices.append(price[1])
            price_times.append(price[0])
        self.electricity_prices = np.array(prices).astype(np.float64)
        self.price_times = np.array(price_times).astype(np.float64)

get_klas ¶

get_klas(step_simtime, s_nh_eff)

Returns the KLA values for the reactor compartments based on electricity prices and ammonia concentration.

Parameters:

Name	Type	Description	Default
`step_simtime`	`int`	Current timestep in the simtime of the simulation [d].	required
`s_nh_eff`	`float`	Ammonia concentration in the plant effluent [g ⋅ m⁻³].	required

Returns:

Name	Type	Description
`klas`	`ndarray`	KLa values for the reactor compartments [d⁻¹]. [kla_reactor1, kla_reactor2, ...]

Source code in src/bsm2_python/controller.py

def get_klas(self, step_simtime: float, s_nh_eff: np.ndarray):
    """Returns the KLA values for the reactor compartments based on electricity prices and ammonia concentration.

    Parameters
    ----------
    step_simtime : int
        Current timestep in the simtime of the simulation [d].
    s_nh_eff : float
        Ammonia concentration in the plant effluent [g ⋅ m⁻³].

    Returns
    -------
    klas : np.ndarray
        KLa values for the reactor compartments [d⁻¹]. \n
        [kla_reactor1, kla_reactor2, ...]
    """

    # necessary to deal with floating point errors
    eps = 1e-8
    step_day_start = np.where(self.price_times - math.floor(step_simtime + eps) <= 0)[0][-1]
    step_day_end = np.where(self.price_times - math.floor(step_simtime + eps + 1) <= 0)[0][-1]
    steps_day = step_day_end - step_day_start
    step_in_day = np.where(self.price_times - (step_simtime + eps) <= 0)[0][-1] - step_day_start

    # get hours with the highest electricity prices at start of day
    if step_in_day == 0:
        self.is_price_in_percentile.clear()
        electricity_prices_day = self.electricity_prices[step_day_start:step_day_end]
        steps_in_percentile = round(steps_day * (1 - self.price_percentile))
        indices_percentile = np.argpartition(electricity_prices_day, -steps_in_percentile)[-steps_in_percentile:]
        for i, _ in enumerate(electricity_prices_day):
            self.is_price_in_percentile.append(i in indices_percentile)
        # add one more step to the end of the day to avoid index out of bounds
        self.is_price_in_percentile.append(False)

    if self.s_nh_threshold > s_nh_eff and self.is_price_in_percentile[step_in_day]:
        klas = self.klas_init * self.kla_reduction
    else:
        klas = self.klas_init

    return klas